One change
Dreams
api/beta/dreams
Nearest release: v2.1.245, 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.
api/beta/dreams Changed · +228 / -179 lines
### Headers ### Body parameters #### Response (200) ### Query parameters ### Headers #### Response (200) ### Path parameters ### Headers #### Response (200) ### Path parameters ### Headers #### Response (200) ### Path parameters ### Headers #### Response (200) ## Domain types ### Header Parameters ### Body Parameters #### Response ### Query Parameters ### Header Parameters #### Response ### Path Parameters ### Header Parameters #### Response ### Path Parameters ### Header Parameters #### Response ### Path Parameters ### Header Parameters #### Response ## Domain Types
The two sides of this change are too far apart to line up, so this is the differ's own diff of it.
from line 1
---- -title: Dreams -url: https://platform.claude.com/docs/en/api/beta/dreams ---- - # Dreams ## Create a Dream -**post** `/v1/dreams` +**POST** `/v1/dreams` Create a Dream -### Header Parameters +### Headers - `"anthropic-beta": optional array of AnthropicBeta`
from line 84
- `"mid-conversation-tool-changes-2026-07-01"` -### Body Parameters +### Body parameters - `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 106
- `type: "sessions"` - - `"sessions"` - - `model: string or BetaDreamModelConfigParam` Model identifier and configuration applied to every pipeline stage. - `string` - - `BetaDreamModelConfigParam object { id, speed }` + - `BetaDreamModelConfigParam object` Model identifier and configuration applied to every pipeline stage.
from line 120
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast" or null` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 132
- `instructions: optional string or null` + minLength: 1, maxLength: 4096 + - `output_behavior: optional BetaOutputBehavior` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - ### Returns -- `BetaDream object { id, archived_at, created_at, 11 more }` +- `BetaDream object` An asynchronous memory-consolidation job that reads a memory store plus a set of session transcripts and writes consolidated memories into an output memory store — a new store by default, or an existing store chosen via output_behavior. The Dreams API is in research preview: the request and response shapes are volatile and may change without the deprecation period that applies to generally-available endpoints.
from line 166
A timestamp in RFC 3339 format + format: date-time + - `created_at: string` A timestamp in RFC 3339 format + format: date-time + - `ended_at: string or null` A timestamp in RFC 3339 format + format: date-time + - `error: BetaDreamError or null` Failure detail for a Dream whose `status` is `failed`.
from line 190
- `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 208
- `type: "sessions"` - - `"sessions"` - - `instructions: string or null` - `model: BetaDreamModelConfig`
from line 218
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 232
The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - - `outputs: array of BetaDreamOutput` - `memory_store_id: string` - `type: "memory_store"` - - `"memory_store"` - - `session_id: string or null` - `status: BetaDreamStatus`
from line 272
- `type: "dream"` - - `"dream"` - - `usage: BetaDreamUsage` Cumulative token usage for the dream across every pipeline stage.
from line 280
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + ### Example -```http +```bash curl https://api.anthropic.com/v1/dreams \ -H 'Content-Type: application/json' \ -H 'anthropic-version: 2023-06-01' \
from line 319
}' ``` -#### Response +#### Response (200) ```json {
from line 365
## List Dreams -**get** `/v1/dreams` +**GET** `/v1/dreams` List Dreams -### Query Parameters +### Query parameters - `"created_at[gt]": optional string` Return dreams with `created_at` strictly after this timestamp (exclusive lower bound, RFC 3339). Unset applies no lower bound. + format: date-time + - `"created_at[lt]": optional string` Return dreams with `created_at` strictly before this timestamp (exclusive upper bound, RFC 3339). Unset applies no upper bound. + format: date-time + - `include_archived: optional boolean` Query parameter for include_archived
from line 391
Query parameter for limit + format: int32 + - `page: optional string` Query parameter for page
from line 411
- `"canceled"` -### Header Parameters +### Headers - `"anthropic-beta": optional array of AnthropicBeta`
from line 499
A timestamp in RFC 3339 format + format: date-time + - `created_at: string` A timestamp in RFC 3339 format + format: date-time + - `ended_at: string or null` A timestamp in RFC 3339 format + format: date-time + - `error: BetaDreamError or null` Failure detail for a Dream whose `status` is `failed`.
from line 523
- `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 541
- `type: "sessions"` - - `"sessions"` - - `instructions: string or null` - `model: BetaDreamModelConfig`
from line 551
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 565
The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - - `outputs: array of BetaDreamOutput` - `memory_store_id: string` - `type: "memory_store"` - - `"memory_store"` - - `session_id: string or null` - `status: BetaDreamStatus`
from line 605
- `type: "dream"` - - `"dream"` - - `usage: BetaDreamUsage` Cumulative token usage for the dream across every pipeline stage.
from line 613
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + - `next_page: string or null` ### Example -```http +```bash curl https://api.anthropic.com/v1/dreams \ -H 'anthropic-version: 2023-06-01' \ -H 'anthropic-beta: dreaming-2026-04-21' \ -H "X-Api-Key: $ANTHROPIC_API_KEY" ``` -#### Response +#### Response (200) ```json {
from line 695
## Get a Dream -**get** `/v1/dreams/{dream_id}` +**GET** `/v1/dreams/{dream_id}` Get a Dream -### Path Parameters +### Path parameters - `dream_id: string` -### Header Parameters +### Headers - `"anthropic-beta": optional array of AnthropicBeta`
from line 783
### Returns -- `BetaDream object { id, archived_at, created_at, 11 more }` +- `BetaDream object` An asynchronous memory-consolidation job that reads a memory store plus a set of session transcripts and writes consolidated memories into an output memory store — a new store by default, or an existing store chosen via output_behavior. The Dreams API is in research preview: the request and response shapes are volatile and may change without the deprecation period that applies to generally-available endpoints.
from line 793
A timestamp in RFC 3339 format + format: date-time + - `created_at: string` A timestamp in RFC 3339 format + format: date-time + - `ended_at: string or null` A timestamp in RFC 3339 format + format: date-time + - `error: BetaDreamError or null` Failure detail for a Dream whose `status` is `failed`.
from line 817
- `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 835
- `type: "sessions"` - - `"sessions"` - - `instructions: string or null` - `model: BetaDreamModelConfig`
from line 845
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 859
The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - - `outputs: array of BetaDreamOutput` - `memory_store_id: string` - `type: "memory_store"` - - `"memory_store"` - - `session_id: string or null` - `status: BetaDreamStatus`
from line 899
- `type: "dream"` - - `"dream"` - - `usage: BetaDreamUsage` Cumulative token usage for the dream across every pipeline stage.
from line 907
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + ### Example -```http +```bash curl https://api.anthropic.com/v1/dreams/$DREAM_ID \ -H 'anthropic-version: 2023-06-01' \ -H 'anthropic-beta: dreaming-2026-04-21' \ -H "X-Api-Key: $ANTHROPIC_API_KEY" ``` -#### Response +#### Response (200) ```json {
from line 982
## Cancel a Dream -**post** `/v1/dreams/{dream_id}/cancel` +**POST** `/v1/dreams/{dream_id}/cancel` Cancel a Dream -### Path Parameters +### Path parameters - `dream_id: string` -### Header Parameters +### Headers - `"anthropic-beta": optional array of AnthropicBeta`
from line 1070
### Returns -- `BetaDream object { id, archived_at, created_at, 11 more }` +- `BetaDream object` An asynchronous memory-consolidation job that reads a memory store plus a set of session transcripts and writes consolidated memories into an output memory store — a new store by default, or an existing store chosen via output_behavior. The Dreams API is in research preview: the request and response shapes are volatile and may change without the deprecation period that applies to generally-available endpoints.
from line 1080
A timestamp in RFC 3339 format + format: date-time + - `created_at: string` A timestamp in RFC 3339 format + format: date-time + - `ended_at: string or null` A timestamp in RFC 3339 format + format: date-time + - `error: BetaDreamError or null` Failure detail for a Dream whose `status` is `failed`.
from line 1104
- `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 1122
- `type: "sessions"` - - `"sessions"` - - `instructions: string or null` - `model: BetaDreamModelConfig`
from line 1132
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 1146
The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - - `outputs: array of BetaDreamOutput` - `memory_store_id: string` - `type: "memory_store"` - - `"memory_store"` - - `session_id: string or null` - `status: BetaDreamStatus`
from line 1186
- `type: "dream"` - - `"dream"` - - `usage: BetaDreamUsage` Cumulative token usage for the dream across every pipeline stage.
from line 1194
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + ### Example -```http +```bash curl https://api.anthropic.com/v1/dreams/$DREAM_ID/cancel \ -X POST \ -H 'anthropic-version: 2023-06-01' \
from line 1224
-H "X-Api-Key: $ANTHROPIC_API_KEY" ``` -#### Response +#### Response (200) ```json {
from line 1270
## Archive a Dream -**post** `/v1/dreams/{dream_id}/archive` +**POST** `/v1/dreams/{dream_id}/archive` Archive a Dream -### Path Parameters +### Path parameters - `dream_id: string` -### Header Parameters +### Headers - `"anthropic-beta": optional array of AnthropicBeta`
from line 1358
### Returns -- `BetaDream object { id, archived_at, created_at, 11 more }` +- `BetaDream object` An asynchronous memory-consolidation job that reads a memory store plus a set of session transcripts and writes consolidated memories into an output memory store — a new store by default, or an existing store chosen via output_behavior. The Dreams API is in research preview: the request and response shapes are volatile and may change without the deprecation period that applies to generally-available endpoints.
from line 1368
A timestamp in RFC 3339 format + format: date-time + - `created_at: string` A timestamp in RFC 3339 format + format: date-time + - `ended_at: string or null` A timestamp in RFC 3339 format + format: date-time + - `error: BetaDreamError or null` Failure detail for a Dream whose `status` is `failed`.
from line 1392
- `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 1410
- `type: "sessions"` - - `"sessions"` - - `instructions: string or null` - `model: BetaDreamModelConfig`
from line 1420
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 1434
The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - - `outputs: array of BetaDreamOutput` - `memory_store_id: string` - `type: "memory_store"` - - `"memory_store"` - - `session_id: string or null` - `status: BetaDreamStatus`
from line 1474
- `type: "dream"` - - `"dream"` - - `usage: BetaDreamUsage` Cumulative token usage for the dream across every pipeline stage.
from line 1482
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + ### Example -```http +```bash curl https://api.anthropic.com/v1/dreams/$DREAM_ID/archive \ -X POST \ -H 'anthropic-version: 2023-06-01' \
from line 1512
-H "X-Api-Key: $ANTHROPIC_API_KEY" ``` -#### Response +#### Response (200) ```json {
from line 1556
} ``` -## Domain Types +## Domain types ### Beta Dream -- `BetaDream object { id, archived_at, created_at, 11 more }` +- `BetaDream object` An asynchronous memory-consolidation job that reads a memory store plus a set of session transcripts and writes consolidated memories into an output memory store — a new store by default, or an existing store chosen via output_behavior. The Dreams API is in research preview: the request and response shapes are volatile and may change without the deprecation period that applies to generally-available endpoints.
from line 1570
A timestamp in RFC 3339 format + format: date-time + - `created_at: string` A timestamp in RFC 3339 format + format: date-time + - `ended_at: string or null` A timestamp in RFC 3339 format + format: date-time + - `error: BetaDreamError or null` Failure detail for a Dream whose `status` is `failed`.
from line 1594
- `inputs: array of BetaDreamInput` - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 1612
- `type: "sessions"` - - `"sessions"` - - `instructions: string or null` - `model: BetaDreamModelConfig`
from line 1622
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 1636
The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - - `outputs: array of BetaDreamOutput` - `memory_store_id: string` - `type: "memory_store"` - - `"memory_store"` - - `session_id: string or null` - `status: BetaDreamStatus`
from line 1676
- `type: "dream"` - - `"dream"` - - `usage: BetaDreamUsage` Cumulative token usage for the dream across every pipeline stage.
from line 1684
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + ### Beta Dream Error -- `BetaDreamError object { message, type }` +- `BetaDreamError object` Failure detail for a Dream whose `status` is `failed`.
from line 1720
An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - - `BetaDreamMemoryStoreInput object { memory_store_id, type }` + - `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - - - `BetaDreamSessionsInput object { session_ids, type }` + - `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 1738
- `type: "sessions"` - - `"sessions"` - ### Beta Dream Memory Store Input -- `BetaDreamMemoryStoreInput object { memory_store_id, type }` +- `BetaDreamMemoryStoreInput object` An input memory store the dream reads from. The dream never mutates this store unless it is also the destination: with output_behavior {type: "update_existing"} the job consolidates this store in place. - `memory_store_id: string` + minLength: 1 + - `type: "memory_store"` - - `"memory_store"` - ### Beta Dream Memory Store Output -- `BetaDreamMemoryStoreOutput object { memory_store_id, type }` +- `BetaDreamMemoryStoreOutput object` An output memory store the dream writes consolidated memories into.
from line 1760
- `type: "memory_store"` - - `"memory_store"` - ### Beta Dream Model Config -- `BetaDreamModelConfig object { id, speed }` +- `BetaDreamModelConfig object` Model identifier and configuration applied to every pipeline stage. Same wire shape as the Agents API ModelConfig.
from line 1770
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast"` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 1782
### Beta Dream Model Config Param -- `BetaDreamModelConfigParam object { id, speed }` +- `BetaDreamModelConfigParam object` Model identifier and configuration applied to every pipeline stage.
from line 1790
Model identifier, e.g. "claude-opus-5". 1-256 characters. + minLength: 1, maxLength: 256 + - `speed: optional "standard" or "fast" or null` Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
from line 1802
### Beta Dream Output -- `BetaDreamOutput object { memory_store_id, type }` +- `BetaDreamOutput object` An output memory store the dream writes consolidated memories into.
from line 1810
- `type: "memory_store"` - - `"memory_store"` - ### Beta Dream Sessions Input -- `BetaDreamSessionsInput object { session_ids, type }` +- `BetaDreamSessionsInput object` Input session transcripts the dream reads.
from line 1820
- `type: "sessions"` - - `"sessions"` - ### Beta Dream Status - `BetaDreamStatus = "pending" or "running" or "completed" or 2 more`
from line 1838
### Beta Dream Usage -- `BetaDreamUsage object { cache_creation_input_tokens, cache_read_input_tokens, input_tokens, output_tokens }` +- `BetaDreamUsage object` Cumulative token usage for the dream across every pipeline stage.
from line 1846
Total tokens used to create prompt-cache entries (sum of all TTL tiers). + format: int32 + - `cache_read_input_tokens: number` Total tokens read from prompt cache. + format: int32 + - `input_tokens: number` Total uncached input tokens consumed across every pipeline stage. + format: int32 + - `output_tokens: number` Total output tokens generated across every pipeline stage. + format: int32 + ### Beta Output Behavior - `BetaOutputBehavior = BetaOutputBehaviorCreateNew or BetaOutputBehaviorUpdateExisting` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - - `BetaOutputBehaviorCreateNew object { type }` + - `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - - - `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` + - `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - `"update_existing"` - ### Beta Output Behavior Create New -- `BetaOutputBehaviorCreateNew object { type }` +- `BetaOutputBehaviorCreateNew object` The default destination: the job creates a new output memory store as a clone of the memory_store input and writes the consolidated memories into it. The input store is never mutated. - `type: "create_new"` - - `"create_new"` - ### Beta Output Behavior Update Existing -- `BetaOutputBehaviorUpdateExisting object { memory_store_id, type }` +- `BetaOutputBehaviorUpdateExisting object` The job writes the consolidated memories into this existing memory store instead of creating one. In EAP the store must be the job's own memory_store input, so the job consolidates the store in place. - `memory_store_id: string` + minLength: 1 + - `type: "update_existing"` - - - `"update_existing"`