Updated October 29, 2026. Rank says GPT-5.4 mini is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty. This is a same-family comparison, so migration details appear when the source data supports them.
GPT-5.4 mini has the higher public score estimate, 52.98 versus 46.15, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 21 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Code generation, repair, and software-engineering tasks
GPT-5.4 mini
GPT-5.4 mini has the higher public coding point estimate, 35.3 to 30.1, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Tool use, computer use, and multi-step task completion
GPT-5.4 mini
GPT-5.4 mini has the higher public agentic point estimate, 36.4 to 33.7, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
1K fresh input + 500 output tokens
GPT-5.4 nano
GPT-5.4 nano has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
200K cached + 20K fresh input + 10K output tokens
GPT-5.4 nano
GPT-5.4 nano has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
50K fresh input + 3K output tokens
GPT-5.4 nano
GPT-5.4 nano has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
Like-for-like · BenchAlign v5.8
GPT-5.4 mini has the higher coding point estimate. Conditional score ranges do not establish rank confidence.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
1 category rests on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
OSWorld-VerifiedAgentic
Normalized gap 33.1Terminal-Bench 2.0Agentic
Normalized gap 13.7ARC-AGI-2Reasoning
Normalized gap 13.2MMMU-ProMultimodal
Normalized gap 10.5MMLU-Pro (Vals)Knowledge
Normalized gap 7.4Each row shows the public-lane category score for both models: the BenchAlign v5.8 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
| Category | GPT-5.4 mini | GPT-5.4 nano | Basis | Reading |
|---|---|---|---|---|
| Agentic | 36.4Supported · #84/135 | 33.7Supported · #97/135 | Like-for-likeBenchAlign v5.8 lane · 6 vs 6 public rows | GPT-5.4 mini leads · intervals overlap |
| Coding | 35.3Supported · #80/156 | 30.1Supported · #100/156 | Like-for-likeBenchAlign v5.8 lane · 5 vs 4 public rows | GPT-5.4 mini leads · intervals overlap |
| Multimodal | 74.0#29/58 | 52.6#51/58 | Like-for-likeProvisional lane · 1 vs 1 weighted rows | GPT-5.4 mini leads |
| Knowledge | 50.5Supported · #77/179 | 39.4Supported · #117/179 | Like-for-likeBenchAlign v5.8 lane · 5 vs 5 public rows | GPT-5.4 mini leads · intervals overlap |
| Instruction following | 89.9#30/157 | 93.3#12/157 | Directional onlyProvisional lane · 0 vs 0 weighted rows | Directional only |
| Reasoning | 43.7Unranked · 4 rankable rows | 40.6Unranked · 4 rankable rows | Not comparableProvisional lane · 1 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | 55.9Unranked · 2 rankable rows | 52.3Unranked · 2 rankable rows | Not comparableProvisional lane · 2 vs 2 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
1K fresh input + 500 output tokens
GPT-5.4 nano has the lower modeled cost
Costs use the listed standard API rates.
50K fresh input + 3K output tokens
GPT-5.4 nano has the lower modeled cost
Costs use the listed standard API rates.
200K cached + 20K fresh input + 10K output tokens
GPT-5.4 nano has the lower modeled cost
Costs use the listed standard API rates.
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
Maximum documented context; output-token limits may be lower.
GPT-5.4 mini
GPT-5.4 nano
GPT-5.4 mini
gpt-5.4-mini
OpenAI GPT-5.4 mini model documentationGPT-5.4 nano
gpt-5.4-nano
OpenAI GPT-5.4 nano model documentationA missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-5.4 mini
$0.075 per 1M cached input tokens
OpenAI pricingGPT-5.4 nano
$0.02 per 1M cached input tokens
OpenAI pricingGPT-5.4 mini
text, image
OpenAI model catalogGPT-5.4 nano
text, image
OpenAI model catalogGPT-5.4 mini
GPT-5.4 nano
GPT-5.4 mini
Generally Available · OpenAI Responses API
OpenAI model catalogGPT-5.4 nano
Generally Available · OpenAI Responses API
OpenAI model catalogGPT-5.4 mini
Reasoning
GPT-5.4 nano
Reasoning
GPT-5.4 mini
Proprietary
GPT-5.4 nano
Proprietary
GPT-5.4 mini
Proprietary
GPT-5.4 nano
Proprietary
GPT-5.4 mini
2026-03-17
GPT-5.4 nano
2026-03-17
GPT-5.4 mini has the higher public score estimate, 52.98 versus 46.15, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
GPT-5.4 mini has the higher public coding point estimate, 35.3 to 30.1, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
GPT-5.4 mini has the higher public agentic tasks point estimate, 36.4 to 33.7, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
For the stated presets, chat costs $0.003 on GPT-5.4 mini and $0.00082 on GPT-5.4 nano; repository review costs $0.051 and $0.01375; the cache-heavy agent loop costs $0.075 and $0.0205. Costs use the listed standard API rates.
Both models list the same context window, 400K.
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Terminal-Bench 2.0
Shared sourceGPT-5.4 mini leads this result
OSWorld-Verified
Shared sourceGPT-5.4 mini leads this result
MCP Atlas
Shared sourceGPT-5.4 mini leads this result
Toolathlon
Shared sourceGPT-5.4 mini leads this result
τ²-bench results
Shared sourceGPT-5.4 mini leads this result
Terminal-Bench 2.1 (Vals)
GPT-5.4 mini leads this result
Terminal-Bench 2.0
Shared sourceGPT-5.4 mini leads this result
Vibe Code Bench
Shared sourceGPT-5.4 mini leads this result
FrontierCode 1.1 Main
Not directly comparable
LiveCodeBench (Vals)
GPT-5.4 nano leads this result
SWE-bench (Vals)
GPT-5.4 mini leads this result
ARC-AGI-1
Shared sourceGPT-5.4 mini leads this result
ARC-AGI-2
Shared sourceGPT-5.4 mini leads this result
MMMU-Pro
Shared sourceGPT-5.4 mini leads this result
MMMU-Pro w/ Python
Shared sourceGPT-5.4 mini leads this result
ExtractBench (LlamaIndex)
Not directly comparable
ParseBench
Not directly comparable
GPQA
Shared sourceGPT-5.4 mini leads this result
GPT-5.4 mini leads this result
HLE w/o tools
Shared sourceGPT-5.4 mini leads this result
GPQA Diamond (Vals)
GPT-5.4 mini leads this result
MMLU-Pro (Vals)
GPT-5.4 mini leads this result
FrontierMath v2 (Tiers 1-3)
Shared sourceGPT-5.4 mini leads this result
FrontierMath v2 (Tier 4)
Shared sourceGPT-5.4 nano leads this result
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 6,100+ subscribers.
Last updated October 29, 2026
