Agentic
Directional only- GPT-5.4 nano
- 34.6
- Supported · #133/152
- GPT-5 (high)
- 48.8
- Estimated · #61/152
- Basis
- BenchAlign lane · 6 vs 1 public rows
- Reading
- Directional only
GPT-5.4 nano vs GPT-5 (high)
Updated September 10, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
GPT-5.4 nano has the higher public score estimate, 59.57 versus 56.3, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
1 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Share or export
Which one for your work
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.
Chat turn cost
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.
Confidence: listed-rates
Cache-heavy agent loop cost
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. GPT-5 (high) has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback
Show secondary and unsupported calls
Repository review cost
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.
Confidence: listed-rates
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
GPT-5 (high) is not ranked on the public lane for coding, so no winner is named for coding.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
GPT-5 (high) is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Confidence: limited
Long documents
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documented
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
17 GPT-5.4 nano only
1 Shared
1 GPT-5 (high) only
- Shared results
- 1
- GPT-5.4 nano only
- 17
- GPT-5 (high) only
- 1
- Like-for-like categories
- 0 / 8
1 category rests on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign 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.
Coding
Not comparable- GPT-5.4 nano
- 37.1
- Supported · #126/151
- GPT-5 (high)
- Not ranked
- Basis
- BenchAlign lane · 3 vs 1 public rows
- Reading
- Not comparable
Reasoning
Not comparable- GPT-5.4 nano
- 73.7
- Unranked · 2 rankable rows
- GPT-5 (high)
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Knowledge
Not comparable- GPT-5.4 nano
- 47.4
- Supported · #97/183
- GPT-5 (high)
- Not ranked
- Basis
- BenchAlign lane · 5 vs 0 public rows
- Reading
- Not comparable
Math
Not comparable- GPT-5.4 nano
- 43.9
- Unranked · 2 rankable rows
- GPT-5 (high)
- Not ranked
- Basis
- Provisional lane · 2 vs 0 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- GPT-5.4 nano
- Not ranked
- GPT-5 (high)
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- GPT-5.4 nano
- 23.8
- #45/48
- GPT-5 (high)
- Not ranked
- Basis
- Provisional lane · 1 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- GPT-5.4 nano
- 93.2
- #9/123
- GPT-5 (high)
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
| Category | GPT-5.4 nano | GPT-5 (high) | Basis | Reading |
|---|---|---|---|---|
| Agentic | 34.6Supported · #133/152 | 48.8Estimated · #61/152 | Directional onlyBenchAlign lane · 6 vs 1 public rows | Directional only |
| Coding | 37.1Supported · #126/151 | Not ranked | Not comparableBenchAlign lane · 3 vs 1 public rows | Not comparable |
| Reasoning | 73.7Unranked · 2 rankable rows | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Knowledge | 47.4Supported · #97/183 | Not ranked | Not comparableBenchAlign lane · 5 vs 0 public rows | Not comparable |
| Math | 43.9Unranked · 2 rankable rows | Not ranked | Not comparableProvisional lane · 2 vs 0 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multimodal | 23.8#45/48 | Not ranked | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Instruction following | 93.2#9/123 | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
A shared-evidence shape is not available.
BenchLM does not draw a radar or infer missing axes when the matched evidence is too sparse.
What each workload costs
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.
Chat turn
1K fresh input + 500 output tokens
- GPT-5.4 nano
- $0.00082
- Fits in one request
- GPT-5 (high)
- $0.00625
- Fits in one request
GPT-5.4 nano has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- GPT-5.4 nano
- $0.01375
- Fits in one request
- GPT-5 (high)
- $0.0925
- Fits in one request
GPT-5.4 nano has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- GPT-5.4 nano
- $0.0205
- Fits in one request
- GPT-5 (high)
- $0.375
- Fits in one request
- Cached input priced at the published list-input rate
GPT-5.4 nano has the lower modeled cost
GPT-5 (high) has no published cached-input rate, so cached tokens use its listed input rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationGPT-5.4 nanoGPT-5 (high)
Context window
Maximum documented context; output-token limits may be lower.
GPT-5.4 nano
GPT-5 (high)
400K
API model ID
GPT-5.4 nano
gpt-5.4-nano
OpenAI GPT-5.4 nano model documentationGPT-5 (high)
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-5.4 nano
$0.02 per 1M cached input tokens
OpenAI pricingGPT-5 (high)
Not published
Documented inputs
GPT-5.4 nano
text, image
OpenAI model catalogGPT-5 (high)
Not sourced
Documented outputs
GPT-5.4 nano
GPT-5 (high)
Not sourced
Provider availability
GPT-5.4 nano
Generally Available · OpenAI Responses API
OpenAI model catalogGPT-5 (high)
Not sourced
Reasoning profile
GPT-5.4 nano
Reasoning
GPT-5 (high)
Reasoning
Weight access
GPT-5.4 nano
Proprietary
GPT-5 (high)
Proprietary
License
GPT-5.4 nano
Proprietary
GPT-5 (high)
Proprietary
Release date
GPT-5.4 nano
2026-03-17
GPT-5 (high)
2025-08-07
If you already use one of these models
- Deployment change
- Both entries list OpenAI as the provider. Confirm endpoint, model ID, limits, and feature support before switching.
- Quality signal
- GPT-5.4 nano has the higher public score estimate, 59.57 versus 56.3, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.01375 vs $0.0925. Cache-heavy agent loop: $0.0205 vs $0.375.
- Context tradeoff
- Both models list 400K.
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence19 rows
Agentic
Terminal-Bench 2.0
GPT-5.4 nano46.3%SourceGPT-5 (high)—Not directly comparable
OSWorld-Verified
GPT-5.4 nano39%SourceGPT-5 (high)—Not directly comparable
MCP Atlas
GPT-5.4 nano56.1%SourceGPT-5 (high)—Not directly comparable
Toolathlon
GPT-5.4 nano35.5%SourceGPT-5 (high)—Not directly comparable
τ²-bench results
GPT-5.4 nano92.5%SourceGPT-5 (high)—Not directly comparable
Terminal-Bench 2.1 (Vals)
GPT-5.4 nano41.6%SourceGPT-5 (high)—Not directly comparable
JobBench
GPT-5.4 nano—GPT-5 (high)8.5%SourceNot directly comparable
Coding
Vibe Code Bench
Shared sourceGPT-5.4 nano26.10%GPT-5 (high)20.09%GPT-5.4 nano leads this result
LiveCodeBench (Vals)
GPT-5.4 nano84.0%SourceGPT-5 (high)—Not directly comparable
SWE-bench (Vals)
GPT-5.4 nano69.8%SourceGPT-5 (high)—Not directly comparable
Knowledge
GPQA
GPT-5.4 nano82.8%SourceGPT-5 (high)—Not directly comparable
HLE
GPT-5.4 nano37.7%SourceGPT-5 (high)—Not directly comparable
HLE w/o tools
GPT-5.4 nano24.3%SourceGPT-5 (high)—Not directly comparable
GPQA Diamond (Vals)
GPT-5.4 nano77.5%SourceGPT-5 (high)—Not directly comparable
MMLU-Pro (Vals)
GPT-5.4 nano77.2%SourceGPT-5 (high)—Not directly comparable
Math
Multimodal
Frequently asked questions
Which is better, GPT-5.4 nano or GPT-5 (high)?
GPT-5.4 nano has the higher public score estimate, 59.57 versus 56.3, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, GPT-5.4 nano or GPT-5 (high)?
GPT-5 (high) is not ranked on the public lane for coding, so no winner is named for coding.
Which is better for agentic tasks, GPT-5.4 nano or GPT-5 (high)?
GPT-5 (high) scores higher for agentic tasks on the public lane, 48.8 to 34.6. GPT-5 (high) is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which costs less, GPT-5.4 nano or GPT-5 (high)?
For the stated presets, chat costs $0.00082 on GPT-5.4 nano and $0.00625 on GPT-5 (high); repository review costs $0.01375 and $0.0925; the cache-heavy agent loop costs $0.0205 and $0.375. GPT-5 (high) has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, GPT-5.4 nano or GPT-5 (high)?
Both models list the same context window, 400K.
Related comparisons
Last updated September 10, 2026
Watch GPT-5.4 nano vs GPT-5 (high)
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.