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/Google
Google

Google · Gemini 3

Gemini 3.7 Flash

Closed weightsSupersededNewproprietary

Previous Flash release (Aug 2026), still offered but superseded three weeks later by Gemini 3.8 Flash at the same price.

AA Intelligence
39.1#23 of 142
Input / output per 1M
$0.75/ $3.75
Output speed
339tok/s
Context window
1M tokens

Composite indexes

  • Artificial Analysis Intelligence Index#23 of 142
    39independentv4.3.2, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: Claude Opus 5.5 at 58

Coding

  • SciCode#15 of 94
    57.2%independentAA run, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: Claude Opus 5.5 at 66.9%

  • Terminal-Bench 4.0#25 of 88
    13.6%independentAA run (Terminal-Bench 4.0), high thinkingArtificial Analysis: Gemini 3.7 Flash ↗
    11.2%independenthigh thinking, mini-SWE-agent harness, ±2.4Terminal-Bench 4.0 official leaderboard ↗

    Leader: Claude Mythos 5.1 at 60.9%

  • Terminal-Bench 2.x#13 of 103
    85.8%independentAA run (Terminal-Bench 2.1), high thinkingArtificial Analysis: Gemini 3.7 Flash ↗
    85.8%vendorTerminal-Bench 2.1, Terminus 2 harness, self-computedGemini 3.8 Flash model card (3.7 Flash column) ↗

    Leader: Claude Fable 5.1 at 91.4%

Agents and tool use

  • GDPval-AA#30 of 103
    1371independentGDPval-AA v2.1, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: Claude Opus 5.5 at 1846

  • OSWorld-Verified#16 of 16
    50.6%vendorOSWorld 2.0 partial scoreGemini 3.8 Flash model card (3.7 Flash column) ↗

    Leader: Kimi K3 at 84.8%

Reasoning and knowledge

  • GPQA Diamond#7 of 142
    94.5%independentAA run, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: GPT-6 Astra at 96.1%

  • Humanity’s Last Exam#12 of 147
    47.9%independentAA run, no tools, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: Claude Opus 5.5 at 61.4%

  • CritPt#32 of 139
    14.3%independentAA run, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: GPT-5.6 Sol at 32.3%

  • ARC-AGI-2#8 of 25
    84.6%independenthigh thinking; ARC Prize verified, cost $0.25/taskARC Prize leaderboard data (semi-private) ↗

    Leader: GPT-6 Astra at 95%

Vision and long context

  • AA Long Context Reasoning#21 of 139
    81.7%independentAA-LCR v1.1, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: Kimi K3 at 88.7%

  • MMMU-Pro#4 of 72
    85.5%independentAA run, high thinkingArtificial Analysis: Gemini 3.7 Flash ↗

    Leader: Claude Opus 5.5 at 87.7%

Human preference

  • LMArena Text#8 of 54
    1490independentgemini-3.7-flash-high; 5,640 votesLMArena text leaderboard ↗

    Leader: Claude Fable 5 at 1506

  • LMArena WebDev#20 of 54
    1595independentgemini-3.7-flash-high; 7,162 votesLMArena Code Arena | WebDev leaderboard ↗

    Leader: Claude Opus 5.5 at 1818

Other published results

ResultScoreReported bySource
DeepSWE v1.165.3%vendorGemini 3.8 Flash model card (3.7 Flash column) ↗

Specs

Released
August 13, 2026
Parameters
Undisclosed
Size class
Undisclosed
Input
text, image, audio, video
Output
text
Context
1M tokens
Max output
Not published
Cached input
$0.075 / 1M
Blended 70/30
$1.65 / 1M
First token
12.59s
  • Price: Gemini Developer API pricing ↗ Same introductory pricing as 3.8 Flash through Dec 31, 2026 ($1.50/$7.50 from Jan 1, 2027). Still listed as a stable model.
  • Speed: Artificial Analysis: Gemini 3.7 Flash (high thinking) ↗

Will it fit your budget?

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

Plan a stack →

Compare with

  • DeepSeekDeepSeek V4.1 Flash39
  • xAIGrok 4.538.8
  • MetaMuse Spark 1.239.6
  • AnthropicClaude Sonnet 538.2