Reka
Reka Flash 3.1
Reka Flash 3.1 is a Reka model with Not verified context, Not verified in / Not verified out per million tokens, last verified September 2, 2026. Open weights: yes. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.
Reka Flash 3.1 is a closed 21B multimodal reasoning model at a low price.
Catalog record checked September 2, 2026Individual provider fields may change
| Specification | Reka Flash 3.1 |
|---|---|
| Provider | Reka |
| Tier | Balanced |
| Context window | Not verifiedUnverified |
| Max output | Not verifiedUnverified |
| Input / 1M tokens | Not verifiedUnverified |
| Output / 1M tokens | Not verifiedUnverified |
| Weights | Open |
| Parameters | 21B |
| Reasoning levels | low, medium, high |
| Modalities | text |
| License | Apache 2.0 |
| API model id | rekaai/reka-flash-3.1 |
| Released | July 12, 2025 |
Pricing tiers: Open weights (Apache 2.0) on Hugging Face; text-only model card. Context window and hosted rates not officially published (Reka's API Flash tier lists $0.80/$2.00). 21B.
Verified evidence
Published benchmark results
Each result keeps its source and measurement date visible. A missing benchmark is not treated as a zero.
| Benchmark | Score | Measured | Source |
|---|---|---|---|
| Artificial Analysis Intelligence Index | 40 | 2026-08-14 | Artificial Analysis · View source |
Best for
- Multimodal
- Cost-sensitive API
- Reasoning
Watch out
Source receipts
Catalog figures for this model were checked against the following sources.
- Reka — Flash 3.1 (accessed 2026-08-29)
- LLMReference — Reka Flash 3.1 (accessed 2026-08-29)
Common questions
Reka Flash 3.1
Answered from the verified figures on this page rather than general guidance.
What is Reka Flash 3.1 best for?
Reka Flash 3.1 is a balanced tier from Reka. It suits multimodal, cost-sensitive api, reasoning. Closed API; smaller context (98K).
Can I self-host Reka Flash 3.1?
Reka Flash 3.1 publishes open weights, but self-hosting depends on the licence, hardware footprint, quantisation quality, and serving stack. A hosted API is often cheaper until you have measured throughput and concurrency on your own hardware.