Capability
61.1/100
field median 56.2
#55 of 232 ranked models
GPT-5.4 mini
Released Mar 17, 2026 — see all recent releases
Decision reading
GPT-5.4 mini scores 61.1 out of 100 and ranks #55 of 232. This profile shows 19 source-displayable benchmark rows; its strongest eligible category is Multimodal & Grounded at #31. API pricing is $0.75 input and $4.5 output per million tokens, with cached input at $0.075.
Data as of September 10, 2026 · How the score is built
Share or export
Strongest published evidence
Multimodal & Grounded ranks #31. Particularly strong for screenshots, documents, charts, and grounded multimodal workflows.
Validate before choosing
19 published rows leave some tracked benchmark slots empty. Agentic is its lowest eligible category at #119.
Check the published record for the model and serving route you use.
Check this modelDecision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Price
$0.75input / $4.50 output
input median $0.97
cached $0.075 · blended $2.63
Speed
201tok/s
field median 86.5 tok/s
First token 3.85 s
Context
400Ktokens
field median 256,000
Maximum output length is tracked separately
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
GPT-5.4 mini category percentile values
- Agentic22nd percentile
- Coding31st percentile
- ReasoningNot eligible
- Knowledge72nd percentile
- MathNot eligible
- MultilingualNot eligible
- Multimodal36th percentile
- Instruction followingNot eligible
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#119/152
- Coding#104/151
- ReasoningNot ranked
- Knowledge#52/183
- MathNot ranked
- MultilingualNot ranked
- Multimodal#31/48
- Inst. FollowingNot ranked
Top decileTop quartileMid-fieldNot eligible
What it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Current modelGPT-5.4 mini · 61.1 score · $2.63 blended per million tokens
Horizontal: blended price per million tokens, log scale · Vertical: public score
How much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic6/6 verified
- Coding4/4 verified
- ReasoningNot measured
- Knowledge4/5 verified
- Math2/2 verified
- MultilingualNot measured
- Multimodal2/2 verified
- Inst. FollowingNot measured
Verified sourceProvisionalNot measured
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- gpt-5.4-miniOpenAI GPT-5.4 mini model documentation
- Context window
- 400KOpenAI GPT-5.4 mini model documentation
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- text, imageOpenAI model catalog
- Output modalities
- textOpenAI model catalog
- Parameters
- Not disclosed by the provider
- Availability
- OpenAI Responses APIOpenAI model catalog
- Cloud regions
- Not tracked yet
- Lifecycle
- activeOpenAI GPT-5.4 mini model documentation
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.075 per million cached input tokensOpenAI pricing
- Self-host
- Weights are not published
- Rate limits
- Not tracked yet
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #119 of 152Percentile 22ndWeight 22%6 benchmarksVerified | 39.1 | #119 of 152 | 22nd | 22% | 6 benchmarks | Verified |
| CodingRank #104 of 151Percentile 31stWeight 20%4 benchmarksVerified | 42.7 | #104 of 151 | 31st | 20% | 4 benchmarks | Verified |
| ReasoningWeight 17%0 benchmarksNot measured | Not measured | Not ranked | Not available | 17% | 0 benchmarks | Not measured |
| KnowledgeRank #52 of 183Percentile 72ndWeight 12%5 benchmarksMixed sources | 55.6 | #52 of 183 | 72nd | 12% | 5 benchmarks | Mixed sources |
| MathRank Not rankedWeight 5%2 benchmarksVerified | 44.5 | Not ranked | Not available | 5% | 2 benchmarks | Verified |
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | Not ranked | Not available | 7% | 0 benchmarks | Not measured |
| MultimodalRank #31 of 48Percentile 36thWeight 12%2 benchmarksVerified | 57.2 | #31 of 48 | 36th | 12% | 2 benchmarks | Verified |
| Inst. FollowingWeight 5%0 benchmarksNot measured | Not measured | Not ranked | Not available | 5% | 0 benchmarks | Not measured |
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| LiveCodeBench (Vals)LiveCodeBench, Vals AI run | Score81.5% | Versus best verified row Best verified: Claude Fable 5.1 · 90.5% | Gap9 behind | WeightWeighted 15% | |
| Vibe Code BenchVibe Code Bench v1.1 | Score47.97% | Versus best verified row Best verified: Claude Opus 4.7 · 71.00% | Gap23 behind | WeightDisplay only | Benchmark exact |
| FrontierCode 1.1 Main | Score27.0% | Versus best verified row Best verified: Claude Fable 5 · 53.5% | Gap26.5 behind | WeightDisplay only | Benchmark exact |
| SWE-bench (Vals)SWE-bench, Vals AI run | Score73.0% | Versus best verified row Best verified: Claude Opus 5 · 97.0% | Gap24 behind | WeightDisplay only | Verified |
Agentic6 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Terminal-Bench 2.0 | Score60% | Versus best verified row Best verified: GPT-5.6 Sol · 91.9% | Gap31.9 behind | WeightWeighted 30% | Provider exact |
| OSWorld-Verified | Score72.1% | Versus best verified row Best verified: Qwen3.8 Max · 86.1% | Gap14 behind | WeightWeighted 25% | Provider exact |
| MCP Atlas | Score57.7% | Versus best verified row Best verified: Muse Spark 1.1 · 88.1% | Gap30.4 behind | WeightDisplay only | Provider exact |
| Toolathlon | Score42.9% | Versus best verified row Best verified: Muse Spark 1.1 · 75.6% | Gap32.7 behind | WeightDisplay only | Provider exact |
| τ²-bench resultsτ²-Bench Tool-Agent-User Evaluation | Score93.4% | Versus best verified row Best verified: GPT-5.4 · 98.9% | Gap5.5 behind | WeightDisplay only | Provider exact |
| Terminal-Bench 2.1 (Vals)Terminal-Bench 2.1, Vals AI run | Score54.7% | Versus best verified row Best verified: GPT-6 Astra · 87.3% | Gap32.6 behind | WeightDisplay only |
Knowledge5 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HLEHumanity's Last Exam | Score41.5% | Versus best verified row Best verified: Claude Fable 5.1 · 65% | Gap23.5 behind | WeightWeighted 35% | Provider exact |
| HLE w/o toolsHumanity's Last Exam without tools | Score28.2% | Versus best verified row Best verified: Claude Fable 5.1 · 60.9% | Gap32.7 behind | WeightWeighted 10% | Provider exact |
| MMLU-Pro (Vals)MMLU-Pro, Vals AI run | Score84.6% | Versus best verified row Best verified: Claude Fable 5.1 · 92.4% | Gap7.8 behind | WeightWeighted 10% | Verified |
| GPQAGraduate-Level Google-Proof Q&A | Score88% | Versus best verified row Best verified: GPT-6 Astra · 96% | Gap8 behind | WeightWeighted 7% | Reported |
| GPQA Diamond (Vals)GPQA Diamond, Vals AI run | Score83.1% | Versus best verified row Best verified: Gemini 3.1 Pro · 95.5% | Gap12.4 behind | WeightDisplay only |
Math2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| FrontierMath v2 (Tiers 1-3)FrontierMath v2 Tiers 1-3 | Score28.280% | Versus best verified row Best verified: GPT-5.6 Sol · 89.000% | Gap60.7 behind | WeightWeighted 30% | Benchmark exact |
| FrontierMath v2 (Tier 4)FrontierMath v2 Tier 4 | Score2.080% | Versus best verified row Best verified: GPT-6 Astra · 97.600% | Gap95.5 behind | WeightWeighted 10% | Benchmark exact |
Multimodal2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| MMMU-ProMassive Multi-discipline Multimodal Understanding Pro | Score76.6% | Versus best verified row Best verified: Gemini 3.5 Flash · 83.6% | Gap7 behind | WeightWeighted 40% | Provider exact |
| MMMU-Pro w/ PythonMMMU-Pro with Python | Score78% | Versus best verified row Best verified: GPT-5.6 Sol · 84.6% | Gap6.6 behind | WeightDisplay only | Provider exact |
Lineage
The sequence follows explicit supersedes links. A successor's displayed score stays at least 0.1 points above its predecessor; raw benchmark rows do not move. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Aug 7, 2025
GPT-5 miniScore 45.1 · $0.25 / $2
Mar 17, 2026 · you are here
GPT-5.4 miniScore 61.1 · $0.75 / $4.5
Mini
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
GPT-5.4 mini ranks #55 of 232 on the public leaderboard with a score of 61.11/100. Its source-verified position is #42 of 130.
GPT-5.4 mini is a proprietary model with a 400K context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
GPT-5.4 mini sits in the GPT-5.4 family with GPT-5.4, GPT-5.4 nano, GPT-5.4 Pro. Its explicit predecessor is GPT-5 mini. 19 of 435 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Multimodal & Grounded at #31, while its lowest eligible position is Agentic at #119. particularly strong for screenshots, documents, charts, and grounded multimodal workflows.
Radar
Full release historyGPT-5.4 mini release history
GPT-5.4 mini and nano released
OpenAI · Model release
Source confirmed
Radar confirmed these at the source. Use GPT-5.4 mini in your work? Explore Radar to follow supported changes and choose your alerts.
Frequently asked questions
How does GPT-5.4 mini perform overall in AI benchmarks?
GPT-5.4 mini ranks #55 out of 232 models on the public BenchAlign leaderboard, with a score of 61.11/100. Its evidence status is Supported, and this profile shows 19 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is GPT-5.4 mini good for knowledge and understanding?
GPT-5.4 mini ranks #52 out of 183 eligible models for knowledge and understanding, with a public category score of 55.6/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is GPT-5.4 mini good for coding and programming?
GPT-5.4 mini ranks #104 out of 151 eligible models for coding and programming, with a public category score of 42.7/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is GPT-5.4 mini good for mathematics?
GPT-5.4 mini has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is GPT-5.4 mini good for agentic tool use and computer tasks?
GPT-5.4 mini ranks #119 out of 152 eligible models for agentic tool use and computer tasks, with a public category score of 39.1/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is GPT-5.4 mini good for multimodal and grounded tasks?
GPT-5.4 mini ranks #31 out of 48 eligible models for multimodal and grounded tasks, with a public category score of 57.2/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Which sibling models are related to GPT-5.4 mini?
GPT-5.4 mini belongs to the GPT-5.4 family. Related tracked variants include GPT-5.4, GPT-5.4 nano, GPT-5.4 Pro. A sibling link indicates shared lineage or a documented configuration relationship; it does not mean the variants have identical pricing, context limits, benchmark evidence, or deployment behavior. Compare before switching.
Does GPT-5.4 mini have full benchmark coverage on BenchLM?
No. GPT-5.4 mini currently has 35 source-displayable rows across 435 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of GPT-5.4 mini?
GPT-5.4 mini has a documented context window of 400K. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.
Compare GPT-5.4 mini with every tracked model482 comparisons
Last updated September 10, 2026. Runtime fields remain blank until a sourced snapshot exists.
Watch GPT-5.4 mini in the weekly brief
Get one weekly email when material rank, price, availability, or benchmark evidence changes are worth revisiting.
Read a sample issueJoin 2,000+ readers.