Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Reka / Reka Flash

Reka Flash (October 2024)

Updated 21B multimodal release with longer video input, native audio understanding and function calling.

Language models (LLMs)
historicalClosed weightsGeneral purposeMultimodal understanding
Official documentation (opens in a new tab)Reviewed 13 Oct 2026
8 of 17 specification fields collected
Context window128K128,000 tokens
Maximum outputNot collectedAwaiting source review
Input billingNot collectedReka historical API
Output billingNot collectedNo inferred price

Specifications

Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.

Context / input limit
128,000 tokensSource for Context / input limit (opens in a new tab)

Provider states 128K; exact binary-token boundary is not specified.

Maximum output
Not collected
Input modalities
Text, Image, Video, AudioSource for Input modalities (opens in a new tab)
Output modalities
Text, AudioSource for Output modalities (opens in a new tab)

Audio generation was experimental and contact-only; it was not supported by the public API.

Reasoning
Not collected
Tool calling
SupportedSource for Tool calling (opens in a new tab)
Structured output
Not collected
Streaming
Not collected
Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
2024-10-04Source for Release date (opens in a new tab)
Parameters
21 billion parametersSource for Parameters (opens in a new tab)
Architecture
Not collected
Languages
English, Indonesian, Thai, Vietnamese, Tamil, Hindi, Arabic, Japanese, Korean, Spanish, French, ItalianSource for Languages (opens in a new tab)

12 focus languages; pretrained on 32.

Context caching
Not collected
Video input duration
3–5 minutes per input; longer videos described through streaming workflowsSource for Video input duration (opens in a new tab)

Provider range, not an exact hard maximum. Timestamp grounding was not yet supported.

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

Reka historical API

Provider source (opens in a new tab)
API model ID
Not collected

An exact current checkpoint selector is not established by the reviewed documentation.

Access
Confirm release-specific access and deployment terms with Reka; a historical announcement does not establish current public API availability.
Regions
Serving-region and residency guarantees require a release-specific agreement.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Retention, training use, regional processing and enterprise terms have not been summarized for this endpoint.

Reka experimental speech access

Provider source (opens in a new tab)
API model ID
Not collected

An exact current checkpoint selector is not established by the reviewed documentation.

Access
The launch directs users to contact Reka for experimental speech output. No public API selector, voice inventory or current availability established.
Regions
Serving-region and residency guarantees require a release-specific agreement.
Languages
EnglishSource for Languages (opens in a new tab)

Speech input/output worked best in English; not a complete language guarantee.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Retention, training use, regional processing and enterprise terms have not been summarized for this endpoint.

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

Reka historical API

Not verified
How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Not collected
Limits & data boundaries
Not collected

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Reka experimental speech access

Not verified
How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Not collected
Limits & data boundaries
Not collected

No verified cache rate for this endpoint. An empty rate is not a zero charge.

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.

The record behind the model

Why it is included

Release-specific inference from the provider’s proprietary-model documentation, native research/service descriptions and public-weight inventory. No weight release for this exact identity was found. Open Edge 2603, Flash 3/3.1 and Inverse Dynamics Model releases are excluded. Historical family names are not equated with later open checkpoints.

Classification evidence (opens in a new tab)

Lifecycle dates

Release announced4 Oct 2024Source (opens in a new tab)

Sources reviewed