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/OpenAI
OpenAI

OpenAI · GPT-5.4

GPT-5.4 mini

Closed weightsSupersededproprietary

Older small OpenAI model still listed on the API pricing page.

AA Intelligence
24.1#59 of 142
Input / output per 1M
$0.75/ $4.50
Output speed
175tok/s
Context window
400K tokens

Composite indexes

  • Artificial Analysis Intelligence Index#59 of 142
    24independentAA Intelligence Index v4.3.2, xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Opus 5.5 at 58

Coding

  • Terminal-Bench 4.0#39 of 88
    2%independentTerminal-Bench 4.0 (AA harness), xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Mythos 5.1 at 60.9%

  • Terminal-Bench 2.x#51 of 103
    59.2%independentTerminal-Bench 2.1 (AA harness), xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Fable 5.1 at 91.4%

  • SciCode#30 of 94
    52.1%independentxhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Opus 5.5 at 66.9%

Agents and tool use

  • GDPval-AA#53 of 103
    1000independentGDPval-AA, xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Opus 5.5 at 1846

  • τ²-Bench#29 of 88
    83.3%independenttelecom (AA), xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: GLM-5.2 at 99.1%

Reasoning and knowledge

  • Humanity’s Last Exam#60 of 147
    28.1%independentno tools (AA), xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Opus 5.5 at 61.4%

  • GPQA Diamond#47 of 142
    87.5%independentAA, xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: GPT-6 Astra at 96.1%

  • CritPt#39 of 139
    10%independentxhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: GPT-5.6 Sol at 32.3%

  • ARC-AGI-2#20 of 25
    18.9%independentsemi-private, XHighARC Prize leaderboard ↗

    Leader: GPT-6 Astra at 95%

Vision and long context

  • AA Long Context Reasoning#52 of 139
    77%independentxhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Kimi K3 at 88.7%

  • MMMU-Pro#43 of 72
    73.3%independentAA, xhighArtificial Analysis: GPT-5.4 mini (xhigh) ↗

    Leader: Claude Opus 5.5 at 87.7%

Human preference

  • LMArena Text#35 of 54
    1448independentgpt-5.4-mini-high, style control default, vote cutoff 2026-09-13LMArena Text leaderboard ↗

    Leader: Claude Fable 5 at 1506

  • LMArena WebDev#46 of 54
    1397independentgpt-5.4-mini-high, LMArena Code arena (formerly WebDev), vote cutoff 2026-09-23LMArena Code (WebDev) leaderboard ↗

    Leader: Claude Opus 5.5 at 1818

Other published results

ResultScoreReported bySource
AA-Omniscience Index · xhigh-18.9 index (-100 to 100)independentArtificial Analysis: GPT-5.4 mini (xhigh) ↗
AutomationBench-AA (partial score) · xhigh26.2%independentArtificial Analysis: GPT-5.4 mini (xhigh) ↗

Specs

Released
March 17, 2026
Parameters
Undisclosed
Size class
Undisclosed
Input
text, image
Output
text
Context
400K tokens
Max output
128K tokens
Cached input
$0.075 / 1M
Blended 70/30
$1.875 / 1M
First token
118.40s
  • Price: OpenAI API pricing ↗ Standard.
  • Speed: Artificial Analysis: GPT-5.4 mini (xhigh) (median output speed; TTFT is time to first chunk incl. reasoning) ↗

Will it fit your budget?

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

Plan a stack →

Compare with

  • Ant GroupLing-3.0-flash-VL24.6
  • xAIGrok 4.324.9
  • Ant GroupLing 3.0 Flash24.9
  • AlibabaQwen3.7-Plus25