Skip to content

GPT-5.4 nano

OpenAIProprietaryReasoningVerifiedReleased Mar 17, 2026Input: file, image, text
Compare with Ling 3.0 Flash VLSource scores ↗OpenRouter ↗
Overall score
46.9
Rank 115 of 222 · 36.2–57.5
Context window
400K
400,000 tokens
Price per 1M tokens
$0.20 in · $1.25 out
✓ Verified on OpenAI's pricing page
Blended price
$0.46
3:1 input:output tokens
Output speed
—
Not measured yet

Category scores

Scores per capability area with the model's rank among ranked models in that category.

  1. 1Instruction following#1193.3
  2. 2Coding#9866.1
  3. 3Math52.3
  4. 4Multimodal#4952.1
  5. 5Knowledge#12249.4
  6. 6Reasoning40.6
  7. 7Agentic & tool use#8920.0
Higher is better. #N is the model's rank in that category, where it is eligible to be ranked.

Benchmark results (15)

As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.

Agentic & tool use

MCP Atlas56.1#32 of 42
Terminal-Bench 2.046.3#33 of 39
OSWorld-Verified39.0#35 of 35
Toolathlon35.5#20 of 22

Multimodal

MMMU-Pro66.1#42 of 48
ParseBench43.35#35 of 37
ExtractBench (LlamaIndex)(lower is better)74.90#5 of 24
MMMU-Pro w/ Python69.5#9 of 9

Knowledge

GPQA82.8#56 of 117
HLE37.7#32 of 69
HLE w/o tools24.3#38 of 43

Reasoning

ARC-AGI-25.7#41 of 41
ARC-AGI-151.50#36 of 36

Math

FrontierMath v2 (Tiers 1-3)44.912#44 of 64
FrontierMath v2 (Tier 4)12.195#43 of 51

Pricing by source

USD per 1M tokens.

SourceInputOutputCached in
✓ Official (OpenAI)
checked Oct 17, 2026 · gpt-5.4-nano
$0.20$1.25$0.020
BenchLM list price$0.20$1.25$0.020
OpenRouter openai/gpt-5.4-nano$0.20$1.25$0.020

Standard tier, prompts under 272K tokens.

OpenAI's current API pricing page lists gpt-5.4-nano at $0.20 input / $0.02 cached input / $1.25 output per million tokens.

Arena ratings

Human-preference ratings from LMArena. Best-rated variant shown.

Text(high)1372#182 · 62.4K votes
Vision(High)1196#96 · 25.6K votes

Speed

Measured by us through OpenRouter (median of trials).

Not measured yet.

Related models

Same family

Similar score