Skip to contentT/TTokens or Towers

Compare

LeaderboardEvery tracked model, ranked and priced side by sideBenchmarksWhat each test measures and who leads itPricesAPI, media, and hardware rates with a cost calculator

Plan

Stack plannerSix questions to a priced local and hosted stackShortlistOur picks for subscriptions, APIs, and hardwareEnterpriseSeat plans against self-hosting for a whole team

Build

HardwareGPUs and workstations, and which models they fitHostsWhere to rent open-weight inference, and for how muchHarnessesCoding agents and the models they drive

Lab

FindingsResults from our own evaluationsScanAudit a repo’s AI usage in the browserMethodWhere the numbers come from and how we check them
Data checked September 25, 2026
Data checked September 25, 2026
Plan a stack
T/TTokens or Towers

An independent guide to AI models, what they cost, and the hardware to run them. No account, no API key, no sponsors.

Compare

  • Leaderboard
  • Benchmarks
  • Prices

Plan

  • Stack planner
  • Shortlist
  • Enterprise

Build

  • Hardware
  • Hosts
  • Harnesses

Lab

  • Findings
  • Scan
  • Method
Data checked September 25, 2026 · Every number links to its sourceRSS · Source on GitHub
Leaderboard/Reka AI
Reka AI

Reka AI · Reka Flash

Reka Flash 3

Open weightsGenerally availableApache 2.0

Reka's 21B open-weights reasoning model (March 2025), served on Reka's own API.

AA Intelligence
5.6#136 of 142
Input / output per 1M
$0.20/ $0.80
Output speed
—
Context window
128K tokens

Composite indexes

  • Artificial Analysis Intelligence Index#136 of 142
    6independentIntelligence Index v4.3; reasoning; estimatedArtificial Analysis: Reka Flash 3 ↗

    Leader: Claude Opus 5.5 at 58

Coding

  • LiveCodeBench#38 of 56
    43.5%independentreasoningArtificial Analysis: Reka Flash 3 ↗

    Leader: Fugu at 92.9%

Agents and tool use

  • τ²-Bench#88 of 88
    0%independenttelecom; reasoningArtificial Analysis: Reka Flash 3 ↗

    Leader: GLM-5.2 at 99.1%

Reasoning and knowledge

  • GPQA Diamond#121 of 142
    52.9%independentreasoningArtificial Analysis: Reka Flash 3 ↗

    Leader: GPT-6 Astra at 96.1%

  • Humanity’s Last Exam#132 of 147
    4.4%independentno tools; reasoningArtificial Analysis: Reka Flash 3 ↗

    Leader: Claude Opus 5.5 at 61.4%

  • CritPt#134 of 139
    0%independentreasoningArtificial Analysis: Reka Flash 3 ↗

    Leader: GPT-5.6 Sol at 32.3%

Vision and long context

  • AA Long Context Reasoning#137 of 139
    0%independentreasoningArtificial Analysis: Reka Flash 3 ↗

    Leader: Kimi K3 at 88.7%

Other published results

ResultScoreReported bySource
aime-2025 · AIME 2025; reasoning33.7%independentArtificial Analysis: Reka Flash 3 ↗
Terminal-Bench Hard (AA) · reasoning0%independentArtificial Analysis: Reka Flash 3 ↗
IFBench · reasoning30.4%independentArtificial Analysis: Reka Flash 3 ↗
AA-Omniscience Index · reasoning-65.1 indexindependentArtificial Analysis: Reka Flash 3 ↗

Specs

Released
March 10, 2025
Parameters
21B
Size class
10 to 40B
Input
text
Output
text
Context
128K tokens
Max output
Not published
Blended 70/30
$0.38 / 1M
  • Price: Artificial Analysis: Reka Flash 3 ↗ First-party (Reka AI) price as listed by Artificial Analysis.

Will it fit your budget?

The planner prices this model against local hardware and other APIs for your workload.

Plan a stack →

Compare with

  • Ai2Olmo 3 7B Think5.6
  • MistralMinistral 3 8B5.5
  • AI21 LabsJamba Reasoning 3B5.7
  • MicrosoftPhi-4 Multimodal Instruct5.8