Reka Flash 3
Reka's 21B reasoning model for chat and coding, served directly by Reka.
- Reasoning
- Structured Output
Added Oct 15, 2026
Model weightsPricing
Auto routing · per 1M tokens- Input
- $0.11
- Output
- $0.21
- Cache read
- $0.0000
Specifications
- Context window
- 65.5K
- Max output
- 65.5K
Benchmarks
Benchmarks
Sourced from Artificial Analysis.
Intelligence Index
5.6
Agentic work
T²-Bench Telecom (legacy)
Legacy fallback · Conversational AI agents in dual-control scenarios
0.0%
Better than 3% of models compared
Document reasoning
AA-LCR v1.1
Long context reasoning with updated grading
0.0%
Better than 5% of models compared
Reasoning
HLE
Humanity's Last Exam
4.4%
Better than 19% of models compared
IFBench
Instruction-following benchmark
30.4%
Better than 15% of models compared
Coding
Terminal-Bench Hard (legacy)
Legacy fallback · Agentic coding and terminal use
0.0%
Better than 5% of models compared
LiveCodeBench
Contamination-free coding benchmark
43.5%
Better than 51% of models compared
Math
AIME 2025
American Invitational Mathematics Examination 2025
33.7%
Better than 34% of models compared
AIME
American Invitational Mathematics Examination
51.0%
Better than 71% of models compared
Math-500
Diverse mathematical problem solving benchmark
89.3%
Better than 62% of models compared
Knowledge
MMLU-Pro
Professional and academic subject knowledge
66.9%
Better than 29% of models compared
Legacy benchmarks
GPQA Diamond (legacy)
Graduate-level scientific reasoning
52.9%
Better than 27% of models compared
AA-LCR (unversioned / legacy)
Long context reasoning evaluation
0.0%
Better than 5% of models compared
Last updated Oct 15, 2026
Artificial AnalysisProviders
Choose explicit providers for this model. Auto routing remains available as the default option.
Loading provider options…