Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d268Policy30d67, 15 pendingSources1,014

GPT-5 mini

OpenAIReleased Aug 7, 2025Current

Score basis

GPT-5 mini is capable in instruction following and long context; and behind the leaders in multimodal tasks, coding, agentic tasks, and math. Too few results yet to rate reasoning, safety, multilingual tasks, or factuality.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Capable2 capabilities
  1. Instruction Following−21.9%2 of 3
  2. Long Context−22.5%1 of 3
Limited4 capabilities
  1. Multimodal−31.9%1 of 6
  2. Coding−39.8%4 of 10
  3. Agentic−43.8%3 of 7
  4. Math−47.1%2 of 5
Not rated4 capabilities

Too few results yet to rate Reasoning, Safety, Multilingual or Factuality.

Price

$0.25input$2.00outputper million tokens

From Artificial Analysis · 2 providers tracked · All prices

Cheaper than 50% of 331 priced models · 3:1 input-to-output blend, log scale

Evidence

96results on46benchmarks

  • 37 independently verified
  • 49 aggregator
  • 5 vendor-reported
  • 5 cross-referenced

From 22 sources · latest Oct 8, 2026 · How verification works

API features

Tool callingStructured outputsReasoning

As listed by OpenRouter

Research

57 papers reference GPT-5 mini

GPT-5 mini benchmark results

96 results on 46 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Instruction Following

CapableFull instruction following ranking

21.9% behind the leader2 of 3 ranked benchmarks measured

Show 2 more instruction following resultsHide 2 instruction following results

Long Context

CapableFull long context ranking

22.5% behind the leader1 of 3 ranked benchmarks measured

Show 1 more long context resultHide 1 long context result

Multimodal

LimitedFull multimodal ranking

31.9% behind the leader1 of 6 ranked benchmarks measured

Show 3 more multimodal resultsHide 3 multimodal results

Coding

LimitedFull coding ranking

39.8% behind the leader4 of 10 ranked benchmarks measured

Show 6 more coding resultsHide 6 coding results

Agentic

LimitedFull agentic ranking

43.8% behind the leader3 of 7 ranked benchmarks measured

Show 2 more agentic resultsHide 2 agentic results

Math

LimitedFull math ranking

47.1% behind the leader2 of 5 ranked benchmarks measured

Show 2 more math resultsHide 2 math results

Reasoning

Not enough dataFull reasoning ranking

0 of 6 ranked benchmarks measured

Show 13 more reasoning resultsHide 13 reasoning results

Factuality

Not enough dataFull factuality ranking

0 of 4 ranked benchmarks measured

Show 5 more factuality resultsHide 5 factuality results

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 37 more resultsHide 37 results

GPT-5 mini: common questions

Who makes GPT-5 mini?

GPT-5 mini is made by OpenAI.

When was GPT-5 mini released?

GPT-5 mini was released on Aug 7, 2025, according to Artificial Analysis.

What is GPT-5 mini good at?

GPT-5 mini is capable in instruction following and long context; and behind the leaders in multimodal tasks, coding, agentic tasks, and math. Too few results yet to rate reasoning, safety, multilingual tasks, or factuality.

How much does GPT-5 mini cost?

GPT-5 mini costs $0.25 per million input tokens and $2.00 per million output tokens, according to Artificial Analysis. We track its price at 2 providers. At a mix of three input tokens to one output token, it is cheaper than 50% of the 331 priced models we track.

How many benchmarks has GPT-5 mini been tested on?

We track 96 results for GPT-5 mini on 46 benchmarks from 22 sources, 37 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does GPT-5 mini support?

OpenRouter lists tool calling, structured outputs, and reasoning for GPT-5 mini.

About this record

Where GPT-5 mini's numbers come from, and every name it appears under.

Tracked since
May 10, 2026
Newest source mention
Aug 25, 2026

Where the results come from

Verification: 96 scores · 37 independently verified · 49 aggregator-attributed · 5 vendor cross-reference · 5 vendor-reported. How these tiers are assigned

From 22 sources on 11 sites. Artificial Analysis supplies 51 of them; the 37 independently verified results come from 9 sites. Bars are coloured by trust tier.

  • artificialanalysis.ai51
  • arcprize.org8
  • epoch.ai7
  • storage.googleapis.com6
  • api.llm-stats.com5
  • swebench.com5
  • x.ai5
  • raw.githubusercontent.com4
  • matharena.ai3
  • datasets-server.huggingface.co1
  • labs.scale.com1

Also known as

How our sources name GPT-5 mini at each reasoning setting.

SettingShort formAPI id
minimalgpt-5 mini (minimal)gpt-5-mini-minimal
lowgpt-5 mini (low)—
mediumgpt-5 mini (2025-08-07) (medium reasoning) gpt-5 mini (medium) gpt 5 mini (2025-08-07) (medium)gpt-5-mini-medium
highgpt-5 mini (high)GPT-5-mini (high) (Non-Reasoning) gpt-5-mini (high) (reasoning) gpt-5-mini-high
Also listed asgpt-5 mini (2025-08-07)gpt-5-mini-2025-08-07gpt-5-mini-thinkinggpt-5-mini:batchOpenAI gpt-5-mini