Cafiyn Pulse
← Startup Signals
6 min readUpdated 2026-08-31Issue 1

The Assistants API is gone. Your app may not have told you yet.

It shut down on 26 August 2026. The replacement is not a drop-in, and the most common migration mistake changes your model's behaviour without throwing an error.

ByKarthik KumarCafiyn Innovations
At a glance
Sunset date
26 August 2026
Status
Already passed. Calls now fail.
Replacement
Responses API, plus Conversations API if you need persistence
Drop-in?
No. The state model is different.
Most common mistake
Assuming previous_response_id carries your instructions

What actually broke

The Assistants API sunset on 26 August 2026 and is no longer available. If you are reading this because something stopped working, that is the reason, and the date has already passed.

The failure is easy to miss. Assistants-based features are usually background work: a support summariser, a document processor, an onboarding agent. They fail into a queue rather than into a user's face, so the first symptom is often a metric drifting rather than an error page. If you are not certain whether you still call it, grep for the endpoints before you read further. It is faster than reasoning about it.

Why this is not a rename

The Assistants API managed a lot on your behalf: threads, runs, the tool attachment lifecycle, and the polling and orchestration around them. You created a thread, added messages, started a run, and waited.

The Responses API does not do that. Calls are stateless. There is no server-side thread lifecycle, which means your application now owns memory, retries, orchestration, and state. That is more code than a find-and-replace, and it is the part teams underestimate.

The vocabulary changed too, in a way that matters. A thread was a collection of messages, and it could only hold messages. A conversation holds items, and an item can be a message, a tool call, a tool output, or other data. If your code assumed everything in the history was a message, that assumption no longer holds.

The gotcha that does not throw

You have three options for carrying history, and the differences between them are not cosmetic.

The one to be careful with is previous_response_id. It is the closest thing to the old behaviour and therefore the one most people reach for, but it does not carry over the previous response's top-level instructions. If your system prompt lived in instructions and you assumed chaining responses would preserve it, your model is now running without it.

Nothing errors. The call succeeds, you get a plausible-looking response, and your assistant has quietly lost its persona, its constraints, and whatever safety rules you wrote into that field. This is the single most expensive detail in the migration, because it degrades quality silently rather than failing loudly. Resend stable instructions on every request.

  • previous_response_id, where OpenAI manages prior context, but you must resend instructions each time
  • Pass prior output items back in yourself, which gives full control, and the option if you need to trim context to manage cost
  • Conversations API, for when you genuinely need a persistent conversation object

What this changes about cost

Owning your own context window is a billing decision as much as an architectural one. When the platform managed the thread, context growth was somebody else's problem. Now every token of history you replay is a token you pay for on every turn.

That cuts both ways. Teams that migrate carelessly replay the entire history each call and watch their bill climb with conversation length. Teams that use the migration as a reason to trim aggressively often end up cheaper than they were before, because they finally have a place to make that decision.

What to do

  1. 1Grep the codebase for the Assistants endpoints and for previous_response_id. Check background jobs and cron tasks, not just request handlers.
  2. 2If you are already on the Responses API, confirm your instructions are being sent on every call rather than assumed to carry over.
  3. 3Decide deliberately which of the three history strategies each feature uses, rather than defaulting to previous_response_id everywhere.
  4. 4Before you migrate, write down what a conversation costs at turn 5 and at turn 50. If those numbers are far apart, trim context as part of the migration rather than afterwards.
  5. 5Add a synthetic check that exercises one real assistant flow end to end, so the next deprecation surfaces as an alert rather than a drifting metric.

Common questions

Is the Assistants API really gone, or just deprecated?

Gone. It sunset on 26 August 2026 and is no longer available. Deprecation notices ran ahead of that date, but the date has now passed, so this is not something to schedule for next quarter.

Can I migrate by swapping the endpoint?

No. The Assistants API managed threads, runs, and orchestration server-side; the Responses API is stateless and leaves memory, retries, orchestration, and state to your application. Expect to write code, not just change a URL.

What is the difference between a thread and a conversation?

A thread stored messages only. A conversation stores items, which include messages, tool calls, and tool outputs. Code that assumed every history entry was a message needs revisiting.

Why did my assistant get worse after migrating?

Most likely your instructions are no longer being sent. previous_response_id does not carry over the previous response's top-level instructions, and nothing errors when they go missing, so the model simply runs without your system prompt. Resend them on every request.

Sources

Read against the primary documentation rather than secondary coverage. Where a figure comes from a provider's own docs, it is quoted as published on 2026-08-31.

Related

What a chatbot workload actually costs →GPT-5 pricing per token →Check your stack across 10 production categories →

More from Signals

OpenAI removes o1, o3-mini, o4-mini and GPT-4 Turbo on 23 October. The replacements it names run from 5.5 times the price to 97 percent less.
Eleven models and six fine tuned models go dark in four days. Here is OpenAI's own mapping, what each landing costs on a typical month of traffic, and the places where the named replacement is not the cheapest or cannot be used at all.
OpenCode says a breach hit 1.4 percent of Console users who linked an AI provider. Rotate those keys before the root cause is published.
OpenCode has confirmed the breach and has not said how it happened or what was taken. Here is what the vendor said, who is exposed, the order to work in, and how to limit the damage next time.
GitHub Actions workflow lockfiles are in public preview. Here is how to turn them on, check them in CI, and what they do not protect.
A tag can move after you reviewed it. A lockfile makes the runner use the commit you reviewed. Here is the setup, the CI check, the policy, and what a lockfile cannot tell you.
Next.js 16.4.1 and 15.5.28 fix two critical remote code execution flaws. For one of them there is no workaround.
Two of the three fixes were postponed from September. Here is what each flaw needs in order to be reachable, the one workaround that exists, and what the announcement leaves out.
A poisoned npm release that punishes you for revoking the stolen token. The order of your response is the whole defence.
Most incident guidance starts with revoking the leaked token. Here that is the trigger. What StepSecurity found, the order it recommends, and why a valid provenance attestation did not help.
Claude Haiku 5.5 costs a tenth of Haiku 4.5 per token. One token past 100,000 it costs half.
The rate card is a tenth of Haiku 4.5. The footnote is a threshold that moves the price fivefold and that a tokenizer change pushes you toward.
Gemini 4 Argon launched at $2 and $10 and goes to $4 and $20 on 1 December. Price your product on the second number.
Every tier on the pricing page doubles on the same date. What changes, what the number includes, and what to do if you are building on the launch rate.
Claude Haiku 5.5 is not a model id swap. Five things that worked on Haiku 4.5 now return a 400.
The model id is the easy part. The requests that now fail, the response parsing that changes, and the details that fail quietly.
Next.js shipped 16.3.7 and 16.3.8 a day apart. Only one of them is the security release.
Two patch releases a day apart, one of them a decoy if you are not reading carefully. And the 15.5 line does not get the high severity fix at all.
The EU AI output marking rule is not a December deadline. For most teams it started in August.
Almost every summary of this gives one date. There are two, and which one applies to you depends on when you shipped, not on how big you are.
Sonnet 5.5 cut the rate by a third and the token count went up by about 30 percent.
A third off the per-token rate is not a third off the bill when the same text now costs more tokens. Here is the arithmetic, and the five ways your code can break.
Your AI gateway holds every provider key you have. A 9.9 just landed in LiteLLM.
An AI gateway concentrates every upstream key into one process. That is the point of it, and it is why a privilege escalation there is worse than it looks.
The table your coding agent created does not have row level security on it.
Dashboard-created tables get row level security by default now. API-created tables do not, and API is how your coding agent works.
A 9.5 in next/og. Whether it reaches you depends on one line in your OG route.
The advisory rates it 9.5, but it only fires if your OG image route puts attacker-controlled text into SVG. Here is how to find out which side of that line you are on.
OpenAI and Anthropic both cut prices on 22 September. One of them costs you a refactor.
Two frontier price cuts in one day is a good week for anyone paying a token bill. The migration cost is not evenly distributed, and one of the two prices comes with an expiry hedge.
A package install reads your home directory. This week two of them did it on purpose.
The implant did nothing clever. It read the files your shell reads, in the place your credentials live, and sent them somewhere else. That is the whole attack, and it works.
GitHub is switching off SHA-1 SSH keys. Find out now whether yours is one of them.
The keys at risk are the old ones, on the machines nobody logs into, in the pipelines that have run untouched for three years. Those are also the ones that will fail loudest.
Every third-party script tag is a standing grant to run code on your visitors.
A marketing form, a chat widget, an analytics snippet. Each one is permission for someone else's server to execute whatever it sends, on every page it sits on, forever.
1.8 million Android apps were scanned for hardcoded keys. Here is how to check yours.
Finding secrets in shipped apps is now a cheap, automated pipeline. Anything compiled into an APK, an IPA or a JavaScript bundle should be treated as already published.
OpenAI's Agents API will hold your agent state. What breaks if you ever need to leave.
Two major providers moved into the agent state layer in the same week. A managed harness is a real shortcut, as long as you keep a copy of what it holds.
Meta's Muse agent can now check out at your store. What merchants should change.
Agent purchases are arriving through wallets and single-use cards, not through merchants adopting a new protocol. Many merchants will receive them without choosing to.
What Brevo's SSO incident teaches about multi-tenant SAML and hijacked sending accounts.
One incident, two lessons: why email authentication did not help, and the tenant-scoping mistake anyone shipping SSO can make.
Cloudflare now sorts bots into Search, Training and Agent. What it means for your site and your agent.
The same change lands on two audiences. Site owners can now accidentally block search; agent builders will find more doors closed.
OpenAI has announced ten dated deadlines between 23 October 2026 and 1 April 2027. Here is the calendar.
Ten dated deadlines in six months, every one announced in public. The failure mode is never the notice. Updated 19 October with the entries added since September.
Nvidia bought Hugging Face. Your open-weight fallback is now a vendor relationship.
The commitments say nothing breaks. The structural change is that "we can always self-host" is now a dependency on one vendor rather than on a neutral commons.
The Videos API goes away on 24 September and OpenAI names nothing to move to.
Every other deprecation on OpenAI's page offers a migration target. This one does not, which makes it a different kind of problem.
Whisper shuts down on 26 February. The risk is not the deadline.
Five months is a generous window. Whisper is also typically wired into a background job that fails into a queue rather than into a user's face.
One Copilot code review costs 13 requests. Here is the rest of the maths.
Copilot now bills on usage rather than seats. The headline change is simple. The multipliers underneath it are where teams get surprised.
Row-level security is on by default now. That is not the same as being protected.
Lovable fixed the default. The gap that actually leaks data was never the default, it was the policy, and no platform can write that for you.
Your model comparison is measuring the wrong number.
Every price list is per token. Your bill is per finished piece of work. Those two numbers can rank models in opposite orders.
An agent with production log access is a new identity. Treat it like one.
Always-on agents that read production logs are arriving fast. The access-control question they raise is old and well understood, which is the good news.

Open the tool.

Ten production categories, three minutes, and a ranked list of what to fix first.

Check your own stack