19 tracked changes to Anthropic models in the twelve months to 2026-09-24, newest first. Each one cites Anthropic's own announcement and carries one recommended action.
These are the ids and API surfaces Anthropic named in the announcements we track. They are what the entries below are about, and they are not a claim about everything Anthropic ships.
Anthropic launched claude-fable-5-1 and claude-mythos-5-1 with a 1M token context window by default, 128k maximum output tokens and always on adaptive thinking, priced at $10 and $50 per million tokens. The notes state that tool_choice types any and tool are not supported and return a 400 error, and that thinking blocks are preserved only for the same model or a newer one.
Agent: tool_choice types any and tool now return an error, so the way the agent is made to call a tool changes. Arize Phoenix reads tool_path, the ordered tool calls with their arguments, and tool_io, what each tool was called with and what it returned, and warns that a correct looking answer can come from a broken path.
Extraction: Forcing a tool call is one way to fix an extractor's output shape, and that request now fails. DeepEval defines a golden as an input plus the expected result, which becomes a full test case once the application produces the actual output, so the same set reruns once the shape is settled.
Revisit the promptSearch your codebase for tool_choice values of any or tool, since those requests now fail, and check any code that replays thinking blocks across models.
Anthropic released v1.0 of the Python SDK. It requires Python 3.10 or later and removes long deprecated surface, including the legacy Text Completions API and the temperature, top_p and top_k parameters on Messages methods. Anthropic's deprecations page states that passing them on v1.0 and later raises a TypeError.
Generation: This is caller code breaking with no model swap on your side. Anthropic deprecated temperature, top_p and top_k from Claude Opus 4.7 onward, and a non-default value returns a 400 error. It is a parameter deprecation; nothing is published about an effect on output quality.
Plan a migrationSearch your code for temperature, top_p and top_k on Messages calls and remove them, and pin your Python SDK version until that change is made.
Anthropic launched claude-opus-5 at $5 and $25 per million tokens, with a 1M token context window as both the default and the maximum and 128k maximum output tokens. The notes state that thinking is on by default, that disabling thinking is only allowed at effort high or below, which they call a breaking change from Claude Opus 4.8, and that Fast mode was removed for Claude Opus 4.7 so those requests return an error.
Agent: Thinking on by default, and a narrower way to turn it off, change how much the agent works out before it acts. Future AGI describes scoring an agent at four resolutions, end to end, trajectory, step and turn, and writes that a high end to end score can hide bad tool calls, while strong per step scores can still miss that the agent pursued the wrong goal.
Chat assistant: Future AGI recommends refreshing a golden dataset immediately after a model update, monthly by default and weekly during rapid change, sampling 200-500 recent production traces stratified by intent and persona and versioning the result as a new artifact.
Generation: If a model from this family also scores your outputs, the judging side moves too. Future AGI writes that a judge model can move 3-5 points on a fixed rubric in a single month without any change on your side.
Revisit the promptFind any request that disables thinking at effort xhigh or max, since that combination is now rejected, and remove the fast mode flag from Claude Opus 4.7 calls.
Anthropic launched claude-sonnet-5 at an introductory $2 and $10 per million tokens with a 1M token context window and 128k maximum output tokens. The notes state that adaptive thinking is on by default, that manual extended thinking was removed and returns a 400 error, that non default sampling parameters return a 400 error, and that a new tokenizer produces about 30 percent more tokens.
Classification: Non default sampling parameters now return a 400 error, which breaks caller code without any model swap on your side. Anthropic first made this parameter deprecation from Claude Opus 4.7 onward.
RAG: A tokenizer that produces about 30 percent more tokens changes how much retrieved text fits in one prompt. Ragas defines faithfulness as the share of claims in a response that can be inferred from the retrieved context, so a smaller retrieved context can move the score by itself.
Summarisation: The same input now costs more tokens, so what reaches the summariser can change before the model does anything. Galtea warns against updating ground truth to match model behaviour when the summaries come back different.
Revisit the promptRecalculate token budgets and cost estimates against the new tokenizer, and strip manual thinking and non default sampling parameters from requests before they start failing.
Anthropic removed fast mode for Claude Opus 4.6. Requests to claude-opus-4-6 with speed set to fast no longer run at fast speed or premium pricing: they run at standard speed, are billed at standard rates, and do not return an error. The usage.speed field on the response reports the speed used, and Anthropic says to migrate to Claude Opus 4.8 to continue using fast mode.
Plan a migrationA request that asks for fast mode on this id still succeeds, so read usage.speed in your own logs to see which speed you are getting, and move to Claude Opus 4.8 if the feature depends on it.
Anthropic announced the deprecation of Claude Opus 4.1, claude-opus-4-1-20250805, with retirement on the Claude API scheduled for August 5, 2026. Anthropic recommended migrating to Claude Opus 4.8.
Chat assistant: Anthropic notified developers on 5 June 2026 that Claude Opus 4.1 would be retired on 5 August 2026, with Claude Opus 4.8 as the recommended replacement. Nothing is published about how the replacement behaves differently, so a forced swap changes the model behind every answer with no provider figure to compare against.
Plan a migrationThe retirement date has passed, so any call still naming claude-opus-4-1-20250805 fails. Confirm nothing does, and record which id your feature runs on now.
Anthropic launched claude-opus-4-8 with a 1M token context window by default on the Claude API, Amazon Bedrock, Google Cloud and Microsoft Foundry. The notes state that the effort parameter defaults to high, that the minimum cacheable prompt length was reduced to 1,024 tokens, that adaptive thinking triggers reasoning only when needed, and that mid conversation system messages are available without a beta header.
Agent: An effort parameter that defaults to high changes how much happens between the request and the answer. Arize Phoenix writes that end to end evaluation only inspects the two endpoints, the input and the final output, and that reasoning, tool selection and the decision path are where the interesting failures sit.
Chat assistant: Mid conversation system messages change the shape of a multi turn request. LangSmith documents four evaluator types for comparing versions, human review, rule based code functions, LLM as judge in reference free and reference based variants, and pairwise comparison.
Rerun the suiteRun your suite once at the default effort of high before you override it, and re-check cache hit rates now that shorter prompts qualify for caching.
Anthropic retired the 1M token context window beta, context-1m-2025-08-07, for Claude Sonnet 4.5 and Claude Sonnet 4. The beta header now has no effect on these models, and requests exceeding the standard 200k token context window return an error. Anthropic says to migrate to Claude Sonnet 4.6 or Claude Opus 4.6, where the 1M context window is included at standard pricing with no beta header required.
RAG: A smaller usable window changes how much retrieved text reaches the model. Ragas defines faithfulness as the share of claims in a response that can be inferred from the retrieved context, so the scorer reads what was actually passed in, not what was retrieved.
Plan a migrationFind any request that still sends the context-1m-2025-08-07 header to these two ids, then either shorten the input or move the feature to a model where the 1M window is standard.
Anthropic launched claude-opus-4-7 at $5 and $25 per million tokens, the same price as Claude Opus 4.6. The notes state that task budgets launched in beta, that high resolution image input is supported up to 2576 pixels on the long edge, and that an xhigh effort level was added.
Agent: Task budgets and a new effort level change how much the agent does before it answers. Future AGI writes that agent traces in contemporary frameworks are routinely 15-40 spans deep, so the comparison reads the trace rather than the final answer alone.
Extraction: Higher resolution image input means page and screenshot fixtures may be re-encoded, which changes the inputs themselves and not only the model. Braintrust tracks every change to a dataset so experiments can pin to specific versions.
Rerun the suiteRun the suite against claude-opus-4-7 at your current effort level first, then once at xhigh, and record what the extra effort changes before you make it the default.
On April 14, 2026, Anthropic notified developers using Claude Sonnet 4 and Claude Opus 4 of their upcoming retirement on the Claude API. Both models retired on June 15, 2026, with claude-sonnet-4-6 and claude-opus-4-8 listed as the recommended replacements.
Chat assistant: The clock on a retirement belongs to the provider. Anthropic says it gives customers with active deployments at least 60 days' notice before retiring a publicly released model, and its page states that partner platforms such as Amazon Bedrock and Google Cloud set their own schedules.
Summarisation: A recommended replacement writes summaries in its own register. Galtea warns against updating ground truth to match model behaviour, so the expected outputs stay where they are and the gap is what gets reported.
Plan a migrationExport your Claude API usage by model and confirm nothing still calls these two dated ids, since requests to them now return an error.
Anthropic took the 1M token context window out of beta for Claude Opus 4.6 and Claude Sonnet 4.6, at standard pricing and with no beta header required. The notes state that the dedicated 1M rate limits were removed and that the media limit was raised to 600 images or PDF pages per request.
RAG: A longer window changes how much retrieved text you put in front of the model. Ragas defines faithfulness as the share of claims in a response that can be inferred from the retrieved context, so the score follows the context you assemble.
Summarisation: Feeding a whole document instead of chunks changes the input to every case, which makes the old run and the new run two different things unless the inputs are tracked. Braintrust tracks every change to a dataset so experiments can pin to specific versions.
Rerun the suiteRemove the 1M beta header from your clients and rerun any long context case, since the separate rate limit pool that used to back those requests is gone.
Anthropic launched automatic caching for the Messages API, where a single cache_control field caches the last cacheable block and the cache point moves forward as a conversation grows. The same entry retired Claude Sonnet 3.7, claude-3-7-sonnet-20250219, and retired Claude Haiku 3.5, claude-3-5-haiku-20241022, on the Claude API. Anthropic recommends upgrading to Claude Sonnet 4.6 and Claude Haiku 4.5.
RAG: When the generating model is replaced, the answer is scored again for faithfulness, which Ragas defines as the share of claims in a response that can be inferred from the retrieved context. It scores grounding in that context, not factual truth.
Plan a migrationCheck that nothing still names either retired id, since Anthropic says requests to retired models fail, and decide separately whether automatic caching should be switched on for your long conversations.
Anthropic launched claude-sonnet-4-6 at $2 and $10 per million tokens with a 1M token context window in beta and extended thinking support. The notes state that web search and web fetch are free when used with code execution, and that web search, web fetch, programmatic tool calling, code execution, tool search, tool use examples and the memory tool are available without a beta header.
Agent: Tools that no longer need a beta header are tools a request can now reach. Arize Phoenix reconstructs tool_path, the ordered tool calls with their arguments, and tool_io, what each tool was called with and what it returned, and warns that a correct looking answer can come from a broken path.
RAG: Where web search or web fetch supplies the documents, the retrieved context changes with the tool path. Ragas defines faithfulness as the share of claims in a response that can be inferred from the retrieved context, and it scores grounding in that context, not factual truth.
Rerun the suiteRemove the beta headers you no longer need from tool requests and rerun the tool using cases, since header changes can alter which tool path a request takes.
Anthropic launched claude-opus-4-6 at $5 and $25 per million tokens. The notes state that adaptive thinking is recommended and manual thinking is deprecated, that an effort parameter with levels low, medium, high, xhigh and max launched, that a 1M token context window is available in beta with long context pricing above 200k input tokens, and that the media limit rose from 100 to 600 images or PDF pages with the 1M context.
Agent: An effort level replaces a thinking budget you used to set by hand, which changes what happens between the request and the answer. Arize Phoenix writes that end to end evaluation only inspects the two endpoints, the input and the final output, and that reasoning, tool selection and the decision path are where the interesting failures sit.
Generation: Moving from a manual thinking budget to adaptive thinking plus an effort level is a prompt change, and the text comes back different. Galtea warns against updating ground truth to match model behaviour.
Revisit the promptSwitch requests from manual thinking budgets to adaptive thinking plus an effort level, then rerun the cases where you previously tuned the thinking budget by hand.
Anthropic took structured outputs out of beta on the Claude API for Claude Sonnet 4.5, Claude Opus 4.5 and Claude Haiku 4.5. The notes state that the output_format parameter moved to output_config.format, and that structured outputs remain in public beta on Amazon Bedrock and Microsoft Foundry.
Extraction: The parameter that carries your schema moved, so the request changes while the fields you expect do not. DeepEval defines a golden as an input plus the expected result, which becomes a full test case once the application produces the actual output.
Classification: Where labels come back through a structured output, the shape of the response is part of the feature. Galtea recommends a tiered release gate in which blocking failures such as safety or policy violations block a deploy at any rate, while regressions gate on the change in fail rate.
Revisit the promptMove output_format to output_config.format in your request builders and rerun the schema validation cases on both the Claude API and any partner platform you use.
Anthropic launched Claude Opus 4.5, which the release notes describe as its most intelligent model, combining maximum capability with practical performance. The release notes name the family alias claude-opus-4-5. Anthropic's model deprecations table lists the dated snapshot claude-opus-4-5-20251101 as Active, with retirement not sooner than 24 November 2026.
Rerun the suiteA new model is an option rather than an event, unless an alias you already call now resolves to it. Run your own dataset against the model you use today and against this one before deciding to move.
On October 28, 2025, Anthropic notified developers using Claude Sonnet 3.7 of its upcoming retirement on the Claude API. The model retired on February 19, 2026, with claude-sonnet-4-6 listed as the recommended replacement.
Classification: A retired id stops serving, so an unattended classifier fails outright rather than quietly scoring worse. Anthropic says it gives customers with active deployments at least 60 days' notice before retiring a publicly released model, and states that partner platforms set their own schedules.
Generation: Prompts written for the retired model produce different text on the replacement. Galtea warns against updating ground truth to match model behaviour.
Plan a migrationConfirm that nothing still calls claude-3-7-sonnet-20250219, since requests to the retired id now fail, and rerun the prompts that were written for it.
Anthropic launched Claude Haiku 4.5, which the release notes describe as its fastest and most intelligent Haiku model, with near-frontier performance. The release notes name the family alias claude-haiku-4-5. Anthropic's model deprecations table lists the dated snapshot claude-haiku-4-5-20251001 as Active, with retirement not sooner than 15 October 2026.
Rerun the suiteIf you are considering this model to cut cost or latency, the comparison that matters is your own dataset scored on both models, not a published summary of either.
Anthropic launched claude-sonnet-4-5-20250929, which it describes as its best model for complex agents and coding. The notes state that a new stop reason model_context_window_exceeded was added, that the memory tool launched in beta and that context editing launched in beta.
RAG: Context editing and a new stop reason change what actually reaches the model on a long input. Ragas defines faithfulness as the share of claims in a response that can be inferred from the retrieved context, so the score follows whatever survived.
Agent: A memory tool adds calls that the final answer never shows. Arize Phoenix reconstructs tool_path, the ordered tool calls with their arguments, and tool_io, what each tool was called with and what it returned.
Rerun the suiteAdd handling for the model_context_window_exceeded stop reason so long inputs fail visibly, and rerun the long context cases that previously truncated.
Change-log entries restate provider announcements with sources and dates. Impact notes cite public practitioner reports. The site publishes no measurements of its own on any model.
Questions about Anthropic changes
Where do the Anthropic entries come from?
From Anthropic's own pages. Every entry links to the changelog, release note, deprecation table or pricing page that carries the announcement. Nothing here is sourced from coverage or from reports by users.
Does DriftCaught test Anthropic models?
No. This site publishes no measurement of any model. We restate what the provider announced and cite published practitioner guides for what that kind of change typically touches.
What should I do when Anthropic announces something?
Check whether the model id you run is named. If it is, rerun your suite against your dataset and read the result against the thresholds you agreed. Each entry carries one recommended action: rerun, re-pin, re-prompt or migrate.
Is every Anthropic announcement here?
No. The log covers changes to the model families we track for customers, inside a twelve-month window, where the announcement is reachable on Anthropic's own domain. Send us a link to anything missing and we will add it.