Batches changedapi/beta/messages/batches
Nearest release: v2.1.258, published an hour after this site recorded the change. Shown because the two are within 24 hours of each other. Nothing here says the release caused the edit.
Recorded here
Lines+357added
Lines−31removed
From line
18
where the diff opens
First seen
14 Aug 2026
this site's first read of the page
Recorded edits33to this page, all time
The whole hunk
from line 18, old and new numbered
/
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 18
1818
1919 - `string`
2020
21 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 38 more`
21 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
2222
2323 - `"message-batches-2024-09-24"`
2424
from line 102
102102
103103 - `"ce-user-management-2026-07-13"`
104104
105 - `"mid-conversation-output-config-2026-07-01"`
106
107 - `"thinking-binding-controls-2026-08-01"`
108
109 - `"mid-conversation-system-clear-at-2026-08-21"`
110
105111- `"anthropic-user-profile-id": optional string`
106112
107113 The user profile ID to attribute the requests in this batch to. Use when acting on behalf of a party other than your organization. Requires the `user-profiles` beta header. Applies to every request in the batch; an individual request whose `user_profile_id` body field conflicts with this header is errored.
from line 1420
14141420
14151421 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
14161422
1417 - `"claude-sonnet-5" or "claude-fable-5" or "claude-mythos-5" or 12 more`
1423 - `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`
14181424
14191425 The model that will complete your prompt.
14201426
14211427 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
14221428
1429 - `"claude-fable-5-1"`
1430
1431 Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows
1432
1433 - `"claude-mythos-5-1"`
1434
1435 Our most capable model for cybersecurity and biology research, available through trusted access programs
1436
14231437 - `"claude-sonnet-5"`
14241438
14251439 High-performance model for coding and agents
from line 1514
15001514
15011515 - `"system"`
15021516
1517 - `clear_at: optional "next_user_message" or "never" or null`
1518
1519 How long this system message's text stays in front of the model. `"never"` (the default) renders it on every request that includes it. `"next_user_message"` renders it only for the user turn it follows: once a later `role: "user"` message exists in `messages` the message stays in the array (send it unchanged) but is no longer shown to the model. Only permitted on `role: "system"` messages.
1520
1521 - `"next_user_message"`
1522
1523 - `"never"`
1524
1525 - `output_config: optional BetaSystemMessageOutputConfig or null`
1526
1527 Per-message output configuration on a role:"system" input message.
1528
1529 Fields here apply per-turn; `format` remains top-level only. An
1530 empty `{}` is accepted on a message that carries content; a message
1531 with neither content nor output_config fields is rejected.
1532
1533 - `effort: optional "low" or "medium" or "high" or 2 more or null`
1534
1535 All possible effort levels.
1536
1537 - `"low"`
1538
1539 - `"medium"`
1540
1541 - `"high"`
1542
1543 - `"xhigh"`
1544
1545 - `"max"`
1546
15031547 - `model: Model`
15041548
15051549 The model that will complete your prompt.
from line 1846
18021846
18031847 - `type: "enabled"`
18041848
1849 - `block_binding: optional BetaThinkingBlockBinding or null`
1850
1851 Controls for block binding: what happens when a thinking block this
1852 request sends back fails the conversation check. Every field is optional;
1853 an empty object means every default.
1854
1855 - `prefix_mismatch_behavior: optional BetaThinkingPrefixMismatchBehavior or null`
1856
1857 What happens when a thinking block in `messages` fails the conversation
1858 check: it was created in a different conversation, or the messages before
1859 it have changed since. `"error"` (the default) fails the request with a
1860 400 error. `"drop_block"` removes the failing blocks and the request
1861 proceeds; the model no longer sees the dropped reasoning.
1862
1863 - `"error"`
1864
1865 - `"drop_block"`
1866
18051867 - `display: optional "summarized" or "omitted" or "updates" or null`
18061868
18071869 Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.
from line 1882
18201882
18211883 - `type: "adaptive"`
18221884
1885 - `block_binding: optional BetaThinkingBlockBinding or null`
1886
1887 Controls for block binding: what happens when a thinking block this
1888 request sends back fails the conversation check. Every field is optional;
1889 an empty object means every default.
1890
18231891 - `display: optional "summarized" or "omitted" or "updates" or null`
18241892
18251893 Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.
from line 2389
23212389
23222390 - `type: "browser_toolset_20260801"`
23232391
2324 - `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`
2325
2326 - `"direct"`
2327
2328 - `"code_execution_20250825"`
2329
2330 - `"code_execution_20260120"`
2331
2332 - `"code_execution_20260521"`
2333
23342392 - `cache_control: optional BetaCacheControlEphemeral or null`
23352393
23362394 Create a cache control breakpoint at this content block.
from line 3015
29573015
29583016 - `type: "computer_toolset_20260801"`
29593017
2960 - `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`
2961
2962 - `"direct"`
2963
2964 - `"code_execution_20250825"`
2965
2966 - `"code_execution_20260120"`
2967
2968 - `"code_execution_20260521"`
2969
29703018 - `cache_control: optional BetaCacheControlEphemeral or null`
29713019
29723020 Create a cache control breakpoint at this content block.
from line 4153
41054153
41064154 - `string`
41074155
4108 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 38 more`
4156 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
41094157
41104158 - `"message-batches-2024-09-24"`
41114159
from line 4237
41894237
41904238 - `"ce-user-management-2026-07-13"`
41914239
4240 - `"mid-conversation-output-config-2026-07-01"`
4241
4242 - `"thinking-binding-controls-2026-08-01"`
4243
4244 - `"mid-conversation-system-clear-at-2026-08-21"`
4245
41924246### Returns
41934247
41944248- `BetaMessageBatch object`
from line 4419
43654419
43664420 - `string`
43674421
4368 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 38 more`
4422 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
43694423
43704424 - `"message-batches-2024-09-24"`
43714425
from line 4503
44494503
44504504 - `"ce-user-management-2026-07-13"`
44514505
4506 - `"mid-conversation-output-config-2026-07-01"`
4507
4508 - `"thinking-binding-controls-2026-08-01"`
4509
4510 - `"mid-conversation-system-clear-at-2026-08-21"`
4511
44524512### Returns
44534513
44544514- `data: array of BetaMessageBatch`
from line 4694
46344694
46354695 - `string`
46364696
4637 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 38 more`
4697 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
46384698
46394699 - `"message-batches-2024-09-24"`
46404700
from line 4778
47184778
47194779 - `"ce-user-management-2026-07-13"`
47204780
4781 - `"mid-conversation-output-config-2026-07-01"`
4782
4783 - `"thinking-binding-controls-2026-08-01"`
4784
4785 - `"mid-conversation-system-clear-at-2026-08-21"`
4786
47214787### Returns
47224788
47234789- `BetaMessageBatch object`
from line 4951
48854951
48864952 - `string`
48874953
4888 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 38 more`
4954 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
48894955
48904956 - `"message-batches-2024-09-24"`
48914957
from line 5035
49695035
49705036 - `"ce-user-management-2026-07-13"`
49715037
5038 - `"mid-conversation-output-config-2026-07-01"`
5039
5040 - `"thinking-binding-controls-2026-08-01"`
5041
5042 - `"mid-conversation-system-clear-at-2026-08-21"`
5043
49725044### Returns
49735045
49745046- `BetaDeletedMessageBatch object`
from line 5100
50285100
50295101 - `string`
50305102
5031 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 38 more`
5103 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
50325104
50335105 - `"message-batches-2024-09-24"`
50345106
from line 5184
51125184
51135185 - `"ce-user-management-2026-07-13"`
51145186
5187 - `"mid-conversation-output-config-2026-07-01"`
5188
5189 - `"thinking-binding-controls-2026-08-01"`
5190
5191 - `"mid-conversation-system-clear-at-2026-08-21"`
5192
51155193### Returns
51165194
51175195- `BetaMessageBatchIndividualResponse object`
from line 6133
60556133
60566134 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
60576135
6058 - `"claude-sonnet-5" or "claude-fable-5" or "claude-mythos-5" or 12 more`
6136 - `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`
60596137
60606138 The model that will complete your prompt.
60616139
60626140 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
60636141
6142 - `"claude-fable-5-1"`
6143
6144 Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows
6145
6146 - `"claude-mythos-5-1"`
6147
6148 Our most capable model for cybersecurity and biology research, available through trusted access programs
6149
60646150 - `"claude-sonnet-5"`
60656151
60666152 High-performance model for coding and agents
from line 6878
67926878
67936879 - `"fast"`
67946880
6881 - `input_transformations: optional array of BetaThinkingDroppedInputTransformation or null`
6882
6883 Changes the API made to the request's input before showing it to the model:
6884 one entry per change, in request order. Today the only entry type is
6885 `thinking_dropped` — a `thinking`, `redacted_thinking` or `connector_text`
6886 block from the request's `messages` that was removed from the prompt instead
6887 of being shown to the model because it failed a binding check. More entry
6888 types may be added over time; ignore types you do not recognize.
6889
6890 Requires `anthropic-beta: thinking-binding-controls-2026-08-01`. Present on
6891 every such response from a model that supports extended thinking, as `[]`
6892 when nothing was changed; without the beta, blocks are removed all the same
6893 but nothing is reported. Removed blocks contribute nothing to
6894 `usage.input_tokens`. When streaming, the array is final in `message_start`;
6895 the final `message_delta` event carries it only when a server-side model
6896 fallback happened mid-stream, in which case it holds the serving model's
6897 entries and replaces the one in `message_start`.
6898
6899 - `path: string`
6900
6901 Where the removed block was in your request, as `messages.{i}.content.{j}`:
6902 `i` indexes the `messages` array you sent and `j` that message's `content`
6903 array — the same form error messages use.
6904
6905 - `reason: "model_binding_mismatch" or "prefix_binding_mismatch" or "organization_binding_mismatch" or "end_user_binding_mismatch"`
6906
6907 Which binding check removed the block: `model_binding_mismatch` — it was
6908 created by a model whose reasoning the requested model may not read;
6909 `prefix_binding_mismatch` — the conversation before it differs from the
6910 conversation it was created in (the rest of that turn's consecutive thinking
6911 blocks are removed with it, each with this reason);
6912 `organization_binding_mismatch` — it was created under a different
6913 organization (an Anthropic organization, AWS account or Google Cloud project)
6914 and this organization is not one of its additional organizations;
6915 `end_user_binding_mismatch` — it was created for a different end user, or
6916 was removed by the consumer-organization binding. A block that would fail
6917 several checks reports one reason, in this order of precedence:
6918 `organization_binding_mismatch`, `end_user_binding_mismatch`,
6919 `model_binding_mismatch`, `prefix_binding_mismatch`.
6920
6921 - `"model_binding_mismatch"`
6922
6923 - `"prefix_binding_mismatch"`
6924
6925 - `"organization_binding_mismatch"`
6926
6927 - `"end_user_binding_mismatch"`
6928
6929 - `type: "thinking_dropped"`
6930
6931 Always `thinking_dropped` for this entry type.
6932
6933 default: thinking_dropped
6934
67956935 - `type: "succeeded"`
67966936
67976937 default: succeeded
from line 8258
81188258
81198259 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
81208260
8121 - `"claude-sonnet-5" or "claude-fable-5" or "claude-mythos-5" or 12 more`
8261 - `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`
81228262
81238263 The model that will complete your prompt.
81248264
81258265 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
81268266
8267 - `"claude-fable-5-1"`
8268
8269 Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows
8270
8271 - `"claude-mythos-5-1"`
8272
8273 Our most capable model for cybersecurity and biology research, available through trusted access programs
8274
81278275 - `"claude-sonnet-5"`
81288276
81298277 High-performance model for coding and agents
from line 8336
81888336
81898337 - `to: BetaFallbackInfo`
81908338
8191 The fallback model producing the content that follows this block. Its `model` is always the canonical id.
8192
8193 - `trigger: BetaFallbackRefusalTrigger`
8194
8195 What caused the `from` model to hand over at this hop.
8196
8197 - `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`
8198
8199 The policy category that triggered a refusal.
8200
8201 - `"cyber"`
8202
8203 The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.
8204
8205 - `"bio"`
8206
8207 The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.
8208
8209 - `"frontier_llm"`
8210
8211 The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.
8212
8213 - `"reasoning_extraction"`
8214
8215 The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).
8216
8217 - `"general_harms"`
8218
8219 The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.
8220
8221 - `type: "refusal"`
8222
8223 default: refusal
8224
8225 - `type: "fallback"`
8226
8227 default: fallback
8228
8229 - `context_management: BetaContextManagementResponse or null`
8230
8231 Context management response.
8232
8233 Information about context management strategies applied during the request.
8234
8235 - `applied_edits: array of BetaClearToolUses20250919EditResponse or BetaClearThinking20251015EditResponse`
8236
8237 List of context management edits that were applied.
8238
8239 - `BetaClearToolUses20250919EditResponse object`
8240
8241 - `cleared_input_tokens: number`
8242
8243 Number of input tokens cleared by this edit.
8244
8245 minimum: 0
8246
8247 - `cleared_tool_uses: number`
8248
8249 Number of tool uses that were cleared.
8250
8251 minimum: 0
8252
8253 - `type: "clear_tool_uses_20250919"`
8254
8255 The type of context management edit applied.
8256
8257 default: clear_tool_uses_20250919
8258
8259 - `BetaClearThinking20251015EditResponse object`
8260
8261 - `cleared_input_tokens: number`
8262
8263 Number of input tokens cleared by this edit.
8264
8265 minimum: 0
8266
8267 - `cleared_thinking_turns: number`
8268
8269 Number of thinking turns that were cleared.
8270
8271 minimum: 0
8272
8273 - `type: "clear_thinking_20251015"`
8274
8275 The type of context management edit applied.
8276
8277 default: clear_thinking_20251015
8278
8279 - `diagnostics: BetaDiagnostics or null`
8280
8281 Response envelope for request-level diagnostics. Present (possibly
8282 null) whenever the caller supplied `diagnostics` on the request.
8283
8284 - `cache_miss_reason: BetaCacheMissModelChanged or BetaCacheMissSystemChanged or BetaCacheMissToolsChanged or 3 more or null`
8285
8286 Explains why the prompt cache could not fully reuse the prefix from the request identified by `diagnostics.previous_message_id`. `null` means diagnosis is still pending — the response was serialized before the background comparison completed.
8287
8288 - `BetaCacheMissModelChanged object`
8289
8290 - `cache_missed_input_tokens: number`
8291
8292 Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.
8293
8294 - `type: "model_changed"`
8295
8296 default: model_changed
8297
8298 - `BetaCacheMissSystemChanged object`
8299
8300 - `cache_missed_input_tokens: number`
8301
8302 Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.
8303
8304 - `type: "system_changed"`
8305
8306 default: system_changed
8307
8308 - `BetaCacheMissToolsChanged object`
8309
8310 - `cache_missed_input_tokens: number`
8311
8312 Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.
8313
8314 - `type: "tools_changed"`
8315
8316 default: tools_changed
8317
8318 - `BetaCacheMissMessagesChanged object`
8319
8320 - `cache_missed_input_tokens: number`
8321
8322 Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.
8323
8324 - `type: "messages_changed"`
8325
8326 default: messages_changed
8327
8328 - `BetaCacheMissPreviousMessageNotFound object`
8329
8330 - `type: "previous_message_not_found"`
8331
8332 default: previous_message_not_found
8333
8334 - `BetaCacheMissUnavailable object`
8335
8336 - `type: "unavailable"`
8337
8338 default: unavailable
8339
8340 - `model: Model`
8341
8342 The model that will complete your prompt.
8343
8344 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
8345
8346 - `role: "assistant"`
8347
8348 Conversational role of the generated message.
8349
8350 This will always be `"assistant"`.
8351
8352 default: assistant
8353
8354 - `stop_details: BetaRefusalStopDetails or null`
8355
8356 Structured information about a refusal.
8357
8358 - `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`
8359
8360 The policy category that triggered a refusal.
8361
8362 - `"cyber"`
8363
8364 The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.
8365
8366 - `"bio"`
8367
8368 The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.
8369
8370 - `"frontier_llm"`
8371
8372 The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.
8373
8374 - `"reasoning_extraction"`
8375
8376 The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking
8339 The fallback model produ
No line in this hunk matches that.