GPT-5.4 Pro
GPT-5.4 Pro is capable in reasoning and math and behind the leaders in factuality. Too few results yet to rate coding, agentic tasks, safety, long context, multimodal tasks, multilingual tasks, or instruction following.
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.
Too few results yet to rate Coding, Agentic, Safety, Long Context, Multimodal, Multilingual or Instruction Following.
Price
$30.00input$180.00outputper million tokens
From Artificial Analysis · 2 providers tracked · All prices
Evidence
19results on17benchmarks
- 11 independently verified
- 1 aggregator
- 7 cross-referenced
From 10 sources · latest Oct 8, 2026 · How verification works
API features
Tool callingStructured outputsReasoning
As listed by OpenRouter
Research
2 papers reference GPT-5.4 ProGPT-5.4 Pro benchmark results
19 results on 17 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
22.7% behind the leader3 of 6 ranked benchmarks measured
- 94.40Jun 21, 2026
- 30.00Oct 8, 2026
- 83.33May 10, 2026
Show 2 more reasoning resultsHide 2 reasoning results
- 83.30Jun 21, 2026
- Humanity's Last Exam42.70May 3, 2026humanitys_last_exam_no_tools
24.5% behind the leader2 of 5 ranked benchmarks measured
- 82.46Sep 21, 2026
- 58.54Sep 21, 2026
31.7% behind the leader2 of 4 ranked benchmarks measured
- 8.30May 2, 2026
- 46.30May 20, 2026
0 of 7 ranked benchmarks measured
- 89.30Jun 21, 2026
0 of 3 ranked benchmarks measured
- 69.23Oct 8, 2026
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- frontiermath_tier_4_v137.50May 20, 2026frontiermath_tier_4
- 94.50May 10, 2026
- vectara_avg_summary_length148.50May 2, 2026Average Summary Length (Words)
- vectara_answer_rate100.00May 2, 2026Answer Rate
- vectara_factual_consistency91.70May 2, 2026Factual Consistency Rate
- 94.50Jun 21, 2026
Show 2 more resultsHide 2 results
- 61.50Jun 21, 2026
- HLE (with tools)58.70Jun 21, 2026Humanity's Last Exam (With tools)
GPT-5.4 Pro: common questions
Who makes GPT-5.4 Pro?
GPT-5.4 Pro is made by OpenAI.
When was GPT-5.4 Pro released?
GPT-5.4 Pro was released on Mar 5, 2026, according to Artificial Analysis.
What is GPT-5.4 Pro good at?
GPT-5.4 Pro is capable in reasoning and math and behind the leaders in factuality. Too few results yet to rate coding, agentic tasks, safety, long context, multimodal tasks, multilingual tasks, or instruction following.
How much does GPT-5.4 Pro cost?
GPT-5.4 Pro costs $30.00 per million input tokens and $180.00 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 costs more than 99% of the 330 priced models we track.
How many benchmarks has GPT-5.4 Pro been tested on?
We track 19 results for GPT-5.4 Pro on 17 benchmarks from 10 sources, 11 of them independently verified. The latest was recorded on Oct 8, 2026.
Which API features does GPT-5.4 Pro support?
OpenRouter lists tool calling, structured outputs, and reasoning for GPT-5.4 Pro.
About this record
Where GPT-5.4 Pro's numbers come from, and every name it appears under.
- Tracked since
- Apr 26, 2026
- Newest source mention
- Aug 29, 2026
Where the results come from
Verification: 19 scores · 11 independently verified · 1 aggregator-attributed · 7 vendor cross-reference. How these tiers are assigned
From 10 sources on 7 sites. www-cdn.anthropic.com supplies 6 of them; the 11 independently verified results come from 4 sites. Bars are coloured by trust tier.
- www-cdn.anthropic.com6
- epoch.ai4
- raw.githubusercontent.com4
- arcprize.org2
- artificialanalysis.ai1
- cdn.sanity.io1
- labs.scale.com1