API changes and deprecations

1. OpenAI removes o1, o3-mini, o4-mini, GPT-4 Turbo and gpt-image-1 on Friday 23 October, with the replacements priced very differently

   23 Oct

   OpenAI's deprecations page lists a 23 October shutdown, announced on 22 April, for o1, o1 pro, o3-mini, o4-mini, GPT-4 (gpt-4-0613), gpt-4-1106-preview, GPT-4 Turbo, GPT-3.5 Turbo 0125, GPT-4.1 nano, gpt-4o-2024-05-13 and gpt-image-1, plus six fine tuned models. It names GPT-5.6 Sol for o1, o1 pro, o3-mini, the GPT-4 models and gpt-4o-2024-05-13, GPT-5.6 Terra for o4-mini and GPT-3.5 Turbo, GPT-5.6 Luna for GPT-4.1 nano, and gpt-image-2.5-sunburst or gpt-image-2.5-flare for gpt-image-1. At the shutdown date the page says a model "will no longer be accessible".

   So what

   Find your calls today, because the failure mode is a forgotten job rather than the main request path. OpenAI documents an optional shutdown_date field on the List models endpoint, so one request lists what your key can still see with a date on it, and a search of your repository and config for the model IDs catches the rest, including per customer settings stored in a database. Then check fine tuning before anything else: the model pages list fine tuning as not supported on Sol, Terra and Luna, so a fine tuned model cannot simply be recreated on the base OpenAI names. The full model by model table, with what each landing costs, is in the analysis.

   Source: OpenAI, API deprecationsIssue 7 · Week of 19 October 2026

2. Google shuts down the Gemini Deep Research agent deep-research-pro-preview-12-2025 on 23 October

   8 Oct

   On 8 October Google's Gemini API release notes deprecated the Deep Research agent deep-research-pro-preview-12-2025 and set its shutdown for 23 October 2026, 15 days later. They name two newer agents: deep-research-preview-04-2026, described as designed for speed and efficiency and suited to streaming back to a client UI, and deep-research-max-preview-04-2026, described as maximum comprehensiveness for automated context gathering and synthesis. The migration step is to change the agent parameter in your interactions.create requests.

   So what

   Swap the agent string this week, and read it as a reminder of what a preview is: OpenAI's own policy says preview models can be retired with notice as short as two weeks, and this one got about that. Decide which of the two replacements fits, since they trade speed for depth, and rerun a sample of your research tasks, because an agent change can alter what comes back and not only the name. Keep preview agents out of any path you cannot change in a day.

   Source: Google, Gemini API release notesIssue 7 · Week of 19 October 2026

3. GitHub Copilot removes Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra on 16 November

   15 Oct

   On 15 October GitHub said it will remove Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra from every Copilot experience on 16 November, including Copilot Chat, inline edits, ask and agent modes, and code completions. The suggested alternatives are Claude Opus 5.5 for both Claude models and GPT-6.1 Sol or GPT-6 Luna for Terra. Opus 4.8 stays for individual subscribers on annual plans. For Copilot Business and Enterprise the alternatives are enabled automatically unless an administrator has turned off the global default or explicitly disabled the model.

   So what

   Check what you have pinned: saved agent configurations, scripts that select a model by name, and organisation model policies. If an admin turned off the global default, the alternatives will not appear on their own, so enable Claude Opus 5.5 and GPT-6.1 Sol in the model policies before 16 November. No action is needed to remove the old models, and the alternatives carry different list prices from the models they replace, so check what your usage costs afterwards.

   Source: GitHub, upcoming Copilot model deprecationsIssue 7 · Week of 19 October 2026

4. Claude Opus 5.5 fast mode is out of research preview at $8 and $40, double the standard price

   15 Oct

   Anthropic's 15 October release notes say fast mode is out of research preview for Claude Opus 5.5: you no longer need to request access, requests no longer need the fast-mode-2026-02-01 beta header (those that still send it keep working), and there is now a default fast mode rate limit in the Start, Build and Scale tiers. Fast mode for Claude Opus 5 and Claude Opus 4.8 stays in research preview. The pricing page lists Opus 5.5 fast mode at $8 input and $40 output per million tokens, against $4 and $20 standard, and says fast mode pricing applies across the full context window, stacks with prompt caching and data residency multipliers, and is not available with the Batch API.

   So what

   Fast mode doubles the price, so spend it where waiting costs you, on interactive paths a person is watching, and not on background work. It cannot be combined with Batch, which is the cheap end of the same dial, so a job that is not latency sensitive should go the other way. Add the model and mode to your cost dashboard as a separate line, because a switch left on in a shared config doubles a bill without changing a single prompt.

   Source: Anthropic, API release notesIssue 7 · Week of 19 October 2026

5. Vercel Token Pass gives a Pro team $80 of monthly open weight model credit for $30, and stops when the credit runs out

   15 Oct

   Vercel Token Pass entered early access for Pro teams on 15 October. For $30 a month a team gets $80 in monthly credit for GLM 5.3 and Kimi K3, with more models to come, through AI Gateway, for use with coding agents or an agent built with the AI SDK. Zero Data Retention is on by default, with inference in the US and EU. The subscription renews monthly, unused credit does not roll over, and when the credit runs out requests stop until the next billing period, with no automatic overage charges.

   So what

   A hard stop is a feature if your fear is a runaway agent bill: the worst month is $30. The trade is that a coding agent that hits the limit mid task simply stops, and only two models are included so far, so check that one of them passes your own evals before you route a team through it. Treat the $80 as a ceiling to try, not a budget, until you have seen a month of real usage.

   Source: Vercel, introducing Token PassIssue 7 · Week of 19 October 2026

6. Unauthenticated Git clones from GitHub can fail during traffic spikes, so give CI a token

   19 Oct

   On 19 October GitHub said that during periods of unusual traffic it may temporarily require authentication for some clones and fetches of public repositories over HTTPS, to protect the availability of github.com. An affected operation fails with a message that authentication is required. Authenticated operations, SSH operations and pushes are unaffected, and GitHub says this is not a general authentication requirement for public repositories. For automated clones it recommends the built in GITHUB_TOKEN in GitHub Actions and, in other CI systems, a GitHub App installation access token, with a personal access token also possible.

   So what

   Find the places your pipeline clones anonymously: dependencies installed from git URLs, submodules, Dockerfile steps and tool installers that fetch from GitHub. Those are the ones that fail in a spike, usually at a bad moment and with an unhelpful error in the middle of a build. Add a token where you can, prefer an app installation token over a personal one outside Actions, and make the failing step retry once after a delay, since GitHub says the protection is lifted again afterwards.

   Source: GitHub, some Git over HTTPS may temporarily require authenticationIssue 7 · Week of 19 October 2026

7. OpenAI deprecated the GPT-4 series on 14 October for removal on 14 January, and the calendar now runs to April 2027

   14 Jan

   OpenAI notified developers on 14 October that gpt-4o-2024-08-06, gpt-4o-2024-11-20, gpt-4o-mini-2024-07-18, gpt-4.1-2025-04-14 and gpt-4.1-mini-2025-04-14 are deprecated and will be removed from the API on 14 January 2027, naming gpt-6-luna as the recommended replacement for all of them. Fine tuned versions "follow the same shutdown schedule as their underlying base models". The 1 October entries are still open: tts-1, tts-1-hd and two gpt-4o-mini-tts snapshots go on 6 January 2027 with gpt-realtime-2.1-mini as the replacement, and gpt-5.3-codex, gpt-5.1 and gpt-5.4-nano go on 1 April 2027. Dates already on the calendar include the GPT-5 and o3 snapshots on 11 December 2026, and the Evals dashboard, Agent Builder and the reusable prompts API on 30 November 2026, with existing evals going read only on 31 October.

   So what

   Search for the five GPT-4 snapshot ids in code, prompt configs, eval harnesses and fine tune pipelines this week, because three months is short for a fine tune migration and a fine tune has to be rebuilt on a new base model rather than swapped. The deprecations page names one replacement for models of very different sizes, so run your own evals before assuming parity, for the 4o and the 4.1 mini alike. GPT-6 Luna lists $0.10 input and $0.50 output per million tokens for short context, so the move is likely to be cheaper per token, but measure it on your prompts. Put 31 October and 30 November on the calendar now if anything you run touches Evals or Agent Builder.

   Source: OpenAI, deprecationsIssue 6 · Week of 12 October 2026

8. Google now routes gemini-3.7-flash and gemini-3.5-flash to their replacements automatically, and a Deep Research agent shuts down on 23 October

   8 Oct

   The Gemini API release notes for 8 October say gemini-3.7-flash is deprecated and replaced by gemini-3.8-flash, and gemini-3.5-flash is deprecated and replaced by gemini-3.6-flash. Neither can be selected in Google AI Studio any more, and API requests to the old ids are "automatically routed to gemini-3.8-flash" and to gemini-3.6-flash respectively. The deep-research-pro-preview-12-2025 agent is deprecated and shuts down on 23 October 2026, with deep-research-preview-04-2026 (speed and efficiency) and deep-research-max-preview-04-2026 (maximum comprehensiveness) named as replacements; the change is the agent parameter in interactions.create requests. On 6 October Google also deprecated gemini-3.1-flash-image in favour of gemini-nano-banana-2.1, with no shutdown date announced.

   So what

   A pinned model string no longer pins anything for those two ids: your requests already run on a different model. If you have evals or regression tests anchored to 3.7 Flash or 3.5 Flash, rerun them against the replacement now and update the strings so your logs say what actually ran. If you use the Deep Research agent preview, you have eight days to change the agent parameter and retest, and the two replacements are positioned differently, so pick on whether you stream to a user interface or want the most thorough result.

   Source: Google, Gemini API release notesIssue 6 · Week of 12 October 2026

9. Managed Agents web tools now obey your network allowlist, and web_fetch only reads URLs it has already seen

   7 Oct

   Anthropic changed how Managed Agents web tools behave on 7 October. With limited networking, the environment allowed_hosts setting now also applies to web_search and web_fetch, a blocked web_fetch returns a url_not_allowed error result, and creating a session fails with a 400 error when an enabled web tool has an allowed_domains entry that is not within allowed_hosts. Also from that date, web_fetch "now fetches only URLs that have already appeared in the session", and otherwise returns a url_not_in_prior_context error result. The two lists match differently: a tool entry covers its subdomains, while an allowed_hosts entry matches one exact host unless it starts with "*.". The documentation example is that the tool entry docs.example.com is not within an allowed_hosts of ["example.com"].

   So what

   Re-run session creation for every agent that uses limited networking, because a definition that worked last week can now fail with a 400. The subtle failure is the match rule: listing example.com in allowed_hosts does not cover docs.example.com, so list the host or use "*.example.com". Then check what your agents do with links. One that used to compose or guess a URL and fetch it will now get url_not_in_prior_context, so put the URLs it should read into the task or have it find them first. Unrestricted networking and self hosted environments do not apply the allowlist to these tools, so a setup that looked fine in a self hosted test can behave differently on the cloud environment.

   Source: Anthropic, API release notesIssue 6 · Week of 12 October 2026

10. Supabase will stop offering four CBC mode TLS 1.2 ciphers on its HTTP APIs from 10 November at the earliest

    10 Nov

    Supabase published a breaking change notice on 12 October saying that from 10 November 2026 at the earliest it will stop offering four CBC mode TLS 1.2 cipher suites on project HTTP APIs, meaning <ref>.supabase.co and custom domains: ECDHE-ECDSA-AES128-SHA256, ECDHE-RSA-AES128-SHA256, ECDHE-ECDSA-AES256-SHA384 and ECDHE-RSA-AES256-SHA384. Auth, the Data API, Storage, Realtime and Edge Functions are affected. Direct Postgres connections, pooler connections and the Management API are not, and TLS 1.3 is unchanged. Supabase says fewer than 0.0001 percent of connections in the past 30 days negotiated these ciphers, that affected customers were emailed on 14 October, and that six GCM and ChaCha20 suites remain.

    So what

    Almost nobody is affected, which is the reason to check rather than skip: the clients that are affected are the ones you will not notice failing, such as embedded devices, kiosks and end of life operating system installs that talk to your project. Find any such client in your fleet and test it against a server limited to the six remaining TLS 1.2 suites or TLS 1.3. This lands alongside the 30 October Data API exposure change we flagged in issue 5, so put both on the same checklist.

    Source: Supabase, changelogIssue 6 · Week of 12 October 2026

11. From April 2027 every App Store submission needs iPhone Duo screenshots

    Apr 2027

    Apple said on 5 October that iPhone Duo will be available to customers from 23 October 2026. Existing apps can be recompiled with Xcode 27.1, which adds development support, and Device Hub shows how an app behaves across all poses and orientations. You can submit iPhone Duo optimized apps in App Store Connect now, and submitting iPhone Duo screenshots adds a badge to your product page. Apple then states: "Starting April 2027, any apps or games submitted will need to include screenshots for iPhone Duo." The post lists no pixel dimensions and points to its screenshot specification page.

    So what

    Two dates for the calendar: 23 October, when real users arrive on the device, and April 2027, when a submission needs these screenshots. Recompile on a branch this month and use Device Hub to find layouts that break in the other poses before a customer does. Then pull the exact dimensions from the screenshot specification page the post links, and add the screenshot task to your release checklist rather than discovering it on the day of a submission.

    Source: Apple, developer newsIssue 6 · Week of 12 October 2026

12. Claude Sonnet 4.5 retires on 30 November, and which platform you are on decides whether that date is yours

    30 Nov

    Anthropic announced on 30 September that claude-sonnet-4-5-20250929 is deprecated, "with retirement on the Claude API scheduled for November 30, 2026", recommending claude-sonnet-5-5 as the replacement. That is roughly 60 days of notice, matching Anthropic's stated policy, and requests to a model past its retirement date fail. The deprecation page adds an important scope note: "The dates on this page apply to Anthropic-operated platforms: the Claude API, Claude Platform on AWS, and Microsoft Foundry. Partner-operated platforms (Amazon Bedrock and Google Cloud) set their own retirement schedules, so a model's lifecycle status and dates can differ."

    So what

    Search your code and prompt configs for the pinned string, including eval harnesses and cron jobs where old model ids survive longest. The AWS wording is a genuine trap: Claude Platform on AWS is Anthropic-operated and on this clock, while Amazon Bedrock is partner-operated and is not, and both run on AWS. Check which one you actually use. Then budget time for the migration rather than a string swap, because Sonnet 5.5 rejects forced tool use, the thinking disabled setting, assistant prefill and non-default sampling parameters, each with a 400.

    Source: Anthropic, model deprecationsIssue 5 · Week of 28 September 2026

13. GitHub starts failing macOS 14 Actions jobs on Monday, and the replacements are all arm64

    5 Oct

    GitHub announced on 1 October that "The macOS 14 runner image will be retired on November 2, 2026", covering the labels macos-14, macos-14-large and macos-14-xlarge. Eight brownout windows precede it, during which jobs using macOS 14 "will temporarily fail", the first running from "October 5, 14:00 UTC to October 6, 00:00 UTC". GitHub also warns it "may reduce macOS 14 runner capacity, which could result in longer queue times" before retirement. The replacements it lists are all described as "macOS arm64 labels": macos-latest, which is macos-26, plus macos-15, macos-latest-xlarge and macos-15-xlarge.

    So what

    The first brownout is Monday afternoon UTC, so this is a today job. Search every workflow and composite action for macos-14 and its large and xlarge variants, including matrix entries. Then treat it as more than a label swap: the replacements are arm64, so any Intel-assuming step, Homebrew path or prebuilt binary in a mac job needs a real test run rather than a rename and a hope.

    Source: GitHub, changelogIssue 5 · Week of 28 September 2026

14. Supabase stops exposing new public tables to the Data API on 30 October, and RLS alone will not save you

    30 Oct

    Supabase's breaking change entry says new tables in the public schema will no longer be exposed to the Data API by default, "enforced on all projects on 2026-10-30". The dashboard control is named "Automatically expose new tables and functions". The entry is explicit that row level security is not the mechanism here: "Without an explicit GRANT, a role will not have access to the created table." It advises bundling GRANT statements alongside the ENABLE ROW LEVEL SECURITY and policy statements in the same migration if you use a custom migration framework.

    So what

    Open each project's Data API settings and note the current state of that toggle, then audit your migration template. A migration that creates a table and only enables RLS plus a policy will produce a table your client SDK cannot read after 30 October, which looks like a mysterious empty result rather than a permissions error. Add the grant to your boilerplate now. If you create tables programmatically at runtime, that code path needs an explicit exposure step.

    Source: Supabase, changelogIssue 5 · Week of 28 September 2026

15. Stripe's new API version removes payment_method_types from Checkout, Payment Intents and Elements

    30 Sep

    Stripe released API version 2026-09-30.endive on 30 September with an unusually broad set of breaking changes. Passing payment_method_types when creating a Checkout Session "now returns a 400 error", and the parameter is removed from Payment Intents and Setup Intents too, with excluded_payment_method_types and a new allowed_payment_method_types named as replacements. In Stripe.js, setting paymentMethodTypes means "Stripe.js now throws an integration error and prevents the Elements instance from being created or updated", and elements.create for the payment request button now throws, with the Express Checkout Element as the replacement. Behaviour also changed where it still works: allowed_payment_method_types "silently filters out incompatible methods" where the old parameter returned an error. Stripe publishes no end-of-life date for older API versions.

    So what

    Nothing breaks on a date here, because Stripe versions are pinned, which is exactly why this is easy to sleep through and expensive later. Grep for payment_method_types across server and client code and for paymentRequestButton or PaymentRequestButtonElement, then confirm your API version is pinned explicitly rather than inherited from the dashboard default. One trap when you do upgrade: snapshot_api_version can only be set when an event destination is created, so moving webhooks to a new version means standing up a parallel endpoint and making your handler idempotent while both are live.

    Source: Stripe, API changelogIssue 5 · Week of 28 September 2026

16. Apple's original Developer ID authority expires on 1 February, and .pkg installers signed by it stop installing

    1 Feb

    Apple posted on 1 October that "The original Developer ID Certification Authority (Sub-CA) expires on February 1, 2027. Certificates issued by this authority will stop working on that date." The consequence is stated plainly: "Starting February 1, 2027, .pkg files signed with an affected certificate will no longer install." The remedy is to generate a replacement from the current authority, Developer ID Certification Authority (G2), which Apple says "is valid until 2031, but the certificates issued by the certificate authority expire annually and must be renewed each year".

    So what

    This hits anyone shipping a signed Mac app, installer or CLI tool outside the App Store. List your Developer ID certificates in the developer portal and work out which came from the original authority rather than G2, then reissue, update the signing identity in your CI secrets and notarisation step, and re-sign anything you still serve for download. Set two reminders: 1 February 2027, and an annual one, because the G2 certificates expire every year.

    Source: Apple, developer newsIssue 5 · Week of 28 September 2026

17. Google Play app registration closes on 30 September, and unregistered apps face global removal

    30 Sep

    Google's policy page sets 30 September 2026 for Play Console app registration, required to meet Android developer verification requirements and Play Console requirements. Google says 99% of apps on Play have been registered automatically, and tells developers to check their Play Console Home page and register any remaining apps to avoid global removal from Google Play. Separately, it says developers can use Play Console to register apps they distribute outside Google Play to ensure those can be installed on certified Android devices.

    So what

    Five days. Open Play Console Home and check every package name you ship, including apps under a second developer account, apps you took over by transfer, and internal testing builds, because those are exactly the listings that fall outside the 99 percent. If you distribute an app outside Play, register it there too or it may stop installing on certified devices.

    Source: Google Play, policy updatesIssue 4 · Week of 21 September 2026

18. Apple's single set of EU business terms starts on 1 October, and you have to actively agree to it

    1 Oct

    From 1 October 2026 apps distributed in the EU fall under one set of terms. The per-install Core Technology Fee is replaced by a 5 percent Core Technology Commission on digital transactions in apps distributed outside the App Store, and the Initial Acquisition Fee and Store Services Fee are eliminated. Apple lists commission of 26 percent on Apple In-App Purchase, or 15 percent for Small Business Program participants, 20 percent and 10 percent for alternative payment processing inside the app, and a Store Services commission of 15 percent and 10 percent on out-of-app offers. Developers select their payment options and must maintain those options for 12 months. Account Holders have to review and agree to the updated licence agreement.

    So what

    Two things get missed here. It is not automatic: an Account Holder has to accept the updated agreement in the developer account, so nothing happens if nobody logs in. And the payment configuration you are live with on 1 October is the one you keep until October 2027, so run the take-rate model before you choose, and check whether Small Business Program enrolment moves you from 26 percent to 15 percent.

    Source: Apple, developer supportIssue 4 · Week of 21 September 2026

19. GitHub put dates on removing ssh-rsa: a 3072-bit minimum from 14 October, brownouts on 4 November and 9 December

    22 Sep

    GitHub published a timeline for removing the ssh-rsa signature type, meaning RSA keys signed with SHA-1, along with the diffie-hellman-group-exchange-sha256 key exchange. All new RSA SSH keys uploaded after 14 October 2026 must be at least 3072 bits, for both signing and authentication. Brownouts follow on 4 November and 9 December 2026. The same 14 October date enables mlkem768x25519-sha256 on github.com. The changelog also prints a final removal date of January 13, 2026, which falls before the brownouts listed above it, so we are not quoting a year for the removal until GitHub corrects the entry.

    So what

    Run ssh -v git@github.com from your CI image and from every deploy box and read the negotiated signature and key exchange algorithms. Audit repository deploy keys and organisation keys for RSA keys under 3072 bits and for anything old enough to be SHA-1 signed, then rotate to ed25519. Treat 4 November as the fire drill: a brownout is the day your pipeline tells you whether you were wrong, at a time you chose.

    Source: GitHub, changelogIssue 4 · Week of 21 September 2026

20. Google's pricing page schedules a doubling of Gemini 3.6, 3.7 and 3.8 Flash prices on 1 January 2027

    1 Jan

    The Gemini Developer API pricing page lists Gemini 3.8 Flash input at $0.75 through 31 December 2026 and $1.50 starting 1 January 2027, with output going from $3.75 to $7.50 per million tokens on the same date. Gemini 3.7 Flash and 3.6 Flash carry identical figures. The same date raises both 3.8 TTS models, both Gemini Robotics ER 2 previews, and context caching, which goes from $0.075 to $0.15 per million tokens on 3.8 Flash. Gemini 3.5 Flash-Lite at $0.30 and $2.50 has no scheduled increase, and neither does any Pro tier.

    So what

    Google does not call this introductory or promotional pricing anywhere on the page. It simply lists two prices with a date between them, which is a scheduled increase by any other name. If a workload runs on a 3.x Flash model, build the 2027 budget on $1.50 and $7.50 now, and benchmark it against Flash-Lite, which carries no scheduled rise. We have logged the date in our price change log.

    Source: Google, Gemini API pricingIssue 4 · Week of 21 September 2026

21. Node 20 is gone from GitHub Actions runners, and the opt-out that let you pin back is gone too

    23 Sep

    GitHub published what it calls the final notification that Node 20 is no longer available on GitHub Actions runners. Runners now use Node 24 for JavaScript actions, maintainers must set an action's runs.using value to node24 and cut a new release, and the temporary ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION opt-out is no longer available. GitHub also flags that Node 24 is incompatible with macOS 13.4 and earlier and has no official ARM32 support, so self-hosted runners on those systems are no longer supported.

    So what

    Search .github and any action.yml you publish for node20 and node16, and move runs.using to node24. There is no longer an escape hatch, so a workflow that depended on the opt-out fails on its next run rather than warning you. If you run self-hosted runners, check the macOS version and the architecture before the next scheduled job rather than after it.

    Source: GitHub, changelogIssue 4 · Week of 21 September 2026

22. Supabase will remove the logs.all Management API endpoint on 23 Sep, and older MCP get_logs setups will break too

    23 Sep

    Supabase will remove the analytics/endpoints/logs.all Management API endpoint on 23 September 2026. Its replacement, analytics/endpoints/logs, accepts only ClickHouse SQL over a single unified logs table. The Supabase MCP server's get_logs tool called logs.all internally, so Supabase says to update to mcp-server-supabase v0.10.0. The dashboard Logs Explorer is not affected.

    So what

    Search your scripts, cron jobs and agent configs for logs.all this week, rewrite those queries in ClickHouse SQL filtering on source_name, and bump mcp-server-supabase for any agent that reads logs. Put 30 October in the calendar too: from then, new public tables on existing projects need an explicit grant before the Data API can see them.

    Source: Supabase, changelogIssue 3 · Week of 14 September 2026

23. DeepSeek cut its Flash model to $0.15 per million input tokens, and old integrations got the cut automatically

    14 Sep

    DeepSeek retired V4 Flash and now lists DeepSeek-V4.1-Flash, API name deepseek-flash, at $0.15 input and $0.60 output off-peak, down from $0.22 and $0.66. Requests to the legacy name deepseek-v4-flash are still accepted, served by V4.1 Flash and billed at the new price. V4 Pro is unchanged.

    So what

    A price cut you receive without a code change, which is unusual. It is also a reminder that a provider can swap the model behind a name you pinned, so if output consistency matters, re-run your evals whenever a pricing page changes. We caught this through our own weekly drift check and logged it in the price change log.

    Source: DeepSeek, API pricingIssue 3 · Week of 14 September 2026

24. Sora 2 and the Videos API are removed on 24 September, with no replacement named

    24 Sep

    OpenAI notified developers on 24 March 2026 that the Videos API and the sora-2 and sora-2-pro models would be removed on 24 September 2026. The deprecation entry names no successor.

    So what

    Every other item on that page offers a migration target. This one does not, which makes it a vendor selection problem rather than a version bump. If video generation is in your product, you have seventeen days to pick a different provider and re-test output quality, not seventeen days to change a model string.

    Source: OpenAI, deprecationsIssue 2 · Week of 7 September 2026

25. Whisper is deprecated, and it is probably in code you have not opened in a year

    26 Aug

    whisper-1, gpt-4o-transcribe, gpt-4o-mini-transcribe and gpt-4o-transcribe-diarize were deprecated on 26 August 2026 and shut down on 26 February 2027. OpenAI names gpt-transcribe and gpt-live-transcribe as replacements.

    So what

    Five months is a generous window. The risk is not the deadline, it is that Whisper is typically wired into a background job written once, and background jobs fail into a queue rather than into a user's face. Grep now, migrate on your own schedule, rather than in February.

    Source: OpenAI, deprecationsIssue 2 · Week of 7 September 2026

26. Agent Builder, the Evals platform and reusable prompts all close on 30 November

    30 Nov

    OpenAI announced on 3 June 2026 that three platform surfaces would shut down on 30 November 2026: the v1/prompts reusable prompts API, the Evals dashboard and API, and Agent Builder.

    So what

    These are the pieces teams build process around rather than product: your evaluation history and your prompt library live here. Export before the date, because a shut-down dashboard takes your baseline measurements with it and you cannot re-run last quarter's evals afterwards.

    Source: OpenAI, deprecationsIssue 2 · Week of 7 September 2026

27. The OpenAI Assistants API is gone

    26 Aug

    The Assistants API shut down on 26 August 2026. OpenAI directs existing users to the Responses API and Conversations API.

    So what

    This date has passed. If anything you run still calls it, it is already broken, and the failure may be sitting in a background job rather than anywhere a user would report it. Check your logs before you read the rest of this list.

    Source: OpenAI API changelogIssue 1 · Week of 31 August 2026

28. GPT-5.4 and 5.4 mini retire in Codex today

    31 Aug

    For users signed in with ChatGPT, GPT-5.4 and GPT-5.4 mini retire in Codex on 31 August 2026. OpenAI recommends GPT-5.6 Terra and GPT-5.6 Luna respectively.

    So what

    If a model name is hardcoded anywhere in your tooling, today is the day it stops resolving. Pin deliberately or read from config, but do not leave it as a literal in a script nobody owns.

    Source: OpenAI API changelogIssue 1 · Week of 31 August 2026

29. GPT-5.6 Sol dropped over 20% for three months

    21 Aug

    OpenAI cut API and credit pricing for GPT-5.6 Sol by more than 20%, described as running for the next three months.

    So what

    A time-boxed cut is a trial balloon, not a new floor. Take the saving, but do not rebuild your unit economics on a price with an expiry date attached.

    Source: OpenAIIssue 1 · Week of 31 August 2026

30. GitHub Copilot moved to usage-based billing

    GitHub announced Copilot is shifting to usage-based billing.

    So what

    Per-seat costs are predictable and usage-based costs are not. If you have agents running unattended, the same change that saves money on light users can produce a genuinely surprising invoice on heavy ones. Set a budget alert now rather than after the first bill.

    Source: GitHub BlogIssue 1 · Week of 31 August 2026

Other topics