reka · Open weights
Reka Flash 3
Reka's 21B reasoning model for chat and coding, served directly by Reka.
- Reasoning
- Structured output
- Input
- Text
- Output
- Text
- Discovery tier
- Open weights
Pricing
USD per 1M tokens- Input
- $0.11
- Output
- $0.21
Reference prices as published by the model provider. API example
Specifications
- Context window
- 65.5K
- Max output
- 65.5K
- Average output (7 days)
- 2.7K tokens
- Longer than 95% of 465 models
Side by side
Reka Flash 3 compared
Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.
| Reka Flash 3rekaThis model | Reka Edge 2603rekaRemove | |
|---|---|---|
| Provider | reka | reka |
| Discovery tier | Open weights | Efficient |
| Added | Oct 15, 2026 | Oct 15, 2026 |
| Limits | ||
| Context window | 65.54KBest | 16.38K |
| Max output | 65.54KBest | 16.38K |
| Average output | 2.7K tokens | 30 tokens |
| Pricing, USD per 1M tokens | ||
| Input | $0.11 | $0.11 |
| Output | $0.21 | $0.11Best |
| Benchmark indexes | ||
| Intelligence Index | 5.6 | — |
| Math Index | 33.7 | — |
| Capabilities | ||
| Reasoning | Yes | No |
| Tools | No | Yes |
| Vision | No | Yes |
| Structured output | Yes | Yes |
| Reasoning effort | — | — |
Public evaluations
Benchmarks
Independent evaluation results, with how each score ranks among the models in the Noumen catalog.
- Reka Flash 3
- Reka Edge 2603
Intelligence Index
- Reka Flash 35.6
- Reka Edge 2603—
Reka Flash 3: Better than 2% of 326 scored models
Indexes
Math Index
Composite of the math evaluations
- Reka Flash 3: 33.7
- Reka Edge 2603: —
Document reasoning
AA-LCRv1.1
Reasoning across long documents
- Reka Flash 3: 0.0%
- Reka Edge 2603: —
Reasoning and knowledge
Humanity’s Last Exam
Expert-level questions across academic fields
- Reka Flash 3: 4.4%
- Reka Edge 2603: —
GPQA Diamond
Graduate-level science questions
- Reka Flash 3: 52.9%
- Reka Edge 2603: —
MMLU-Pro
Multi-subject knowledge and reasoning
- Reka Flash 3: 66.9%
- Reka Edge 2603: —
IFBench
Following precise output instructions
- Reka Flash 3: 30.4%
- Reka Edge 2603: —
Coding
LiveCodeBench
Recent competitive programming problems
- Reka Flash 3: 43.5%
- Reka Edge 2603: —
Math
AIME 2025
Competition mathematics, 2025 problems
- Reka Flash 3: 33.7%
- Reka Edge 2603: —
AIME
Competition mathematics
- Reka Flash 3: 51.0%
- Reka Edge 2603: —
MATH-500
Competition math problems
- Reka Flash 3: 89.3%
- Reka Edge 2603: —
Legacy evaluations
Terminal-Bench Hard
Earlier terminal task suite
- Reka Flash 3: 0.0%
- Reka Edge 2603: —
τ²-Bench Telecom
Tool use in a customer-support simulation
- Reka Flash 3: 0.0%
- Reka Edge 2603: —
AA-LCR (unversioned)
Long-context reasoning, earlier grading
- Reka Flash 3: 0.0%
- Reka Edge 2603: —
Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis
Developers
Use it through the API
Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.
curl https://www.noumen.xyz/api/v1/chat/completions \
-H "Authorization: Bearer $NOUMEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "reka/reka-flash-3",
"messages": [{ "role": "user", "content": "Explain this stack trace." }]
}'import os
from openai import OpenAI
client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
model="reka/reka-flash-3",
messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)reka
Related models
Other reka models, starting with the same family.
- Efficient