GPT-5.5

Apr 2026
GPT-5.5
Compare in
on
$1.54per task?
$5 / $30per 1M in / outcache $0.50
103 tok/sfirst token in 21 s?
1.05Mcontext

Rankingat high effort?

Full ranking →
1Gemini 4 Argon1,608.32Fable 5.11,585.63Opus 5.51,583.64Opus 51,573.45Gemini 3.8 Flash1,550.8
···
19GPT-5.6 Sol1,526.620GPT-5.51,525.121Kimi K2.61,520.7

Strong at

S
Reasoning#23 of 334Science#11 of 312Long context#6 of 273Knowledge#27 of 318Vision#12 of 124
A
Tool calling#29 of 242Instructions#39 of 245
B
Coding#39 of 132Agents & work tasks#45 of 175

Alternatives?

ModelScoreCostPer taskWhat changes
CheaperGemini 3.8 Flashhighwins 54%-19%*$1.24Context: 1.05M → 1M+2OpenGLM 5.3maxsame+30%$2.01Context: 1.05M → 1M+5CheaperClaude Opus 4.6highwins 54%-11%$5 / $25Context: 1.05M → 1MCheaperClaude Opus 4.7highsame-11%$5 / $25Context: 1.05M → 1MCheaperQwen 3.7 Max Previewsame-67%$2.50 / $7.50Context: not verified+1

History

17 versions
1,3001,4001,500Jul 24Jan 25Jul 25Jan 26Release dateArena Text · Chinese4o4.1 · Apr5 · Aug5.1 · Nov5.2 · Dec5.4 · Mar5.5
1,3001,4001,50020252026Release dateArena Text · Chinese4o4.1 · Apr5 · Aug5.1 · Nov5.2 · Dec5.4 · Mar5.5

Benchmarks

in Chat & Knowledge
AA Intelligence Index 4.3.237#41
AA Omniscience Index18.8#27
AA Omniscience · Accuracy57#8
AA Omniscience · Hallucination Rate89.1#239
Humanity's Last Exam45#23
GPQA Diamond93.2#11
AA-LCR (long context) 1.184.3#6
IFBench (instructions)71.6#39
CritPt (physics)25.4#19
GDP.pdf · AA all-pass22.6#19
ARC-AGI-1 · public evaluation97.4#15
ARC-AGI-2 · public evaluation82.2#16
ARC-AGI-1 · semi-private94.5#15
ARC-AGI-2 · semi-private83.3#14
Arena Text · overall1,481#19
Arena Text · overall · full cohort1,471.3#20
Arena Text · Chinese1,525.1#20
Arena Text · English1,470.3#27
Arena Text · French1,483#22
Arena Text · German1,459.2#27
Arena Text · instruction following1,478.5#14
Arena Text · Japanese1,498.5#5
Arena Text · Korean1,459.3#11
Arena Text · non-English1,466.9#17
Arena Text · Polish1,464.4#19
Arena Text · Russian1,472.5#21
Arena Text · Spanish1,468.1#22
Mystery Game Puzzles52#10
HealthBench Professional length-adjustedlab51.8
HealthBench length-adjustedlab56.5
HealthBench Hard length-adjustedlab31.5
HealthBench Consensus length-adjustedlab95.6

Watch out 1

Prompts with >272K input tokens›

Above 272K input tokens, input costs 2× and output 1.5× for the full session in Standard, Batch and Flex. Source ↗

FAQ

How much does GPT-5.5 cost?›

$5 per 1M input tokens and $30 per 1M output tokens; cached input $0.50 per 1M. A typical benchmark task costs about $1.54 at high effort.

Is GPT-5.5 free to use?›

There is no free API; the cheapest paid price is $5 per 1M input tokens.

How good is GPT-5.5, and what is it best at?›

#23 of 334 in Chat & Knowledge, #39 of 132 in Coding, #41 of 133 in Agents & Tool Use, #25 of 195 in Writing & Language, #12 of 124 in Vision & Documents. Strongest at: Tool calling, Reasoning, Science, Long context, Knowledge, Instructions, Vision.

Is GPT-5.5 better than GPT-5.4?›

On Arena Text · Chinese, GPT-5.5 scores 1,533.2 vs 1,518.5 — about the same.

How can I use GPT-5.5?›

Through the API of Amazon Bedrock, Azure, OpenAI.

What is the context window of GPT-5.5?›

1.05M tokens, with up to 128K output tokens.

When was GPT-5.5 released? Is there a newer version?›

Released 2026-04-23. It has been superseded by GPT-5.6 Sol.

Is GPT-5.5 open source?›

No, GPT-5.5 is a closed model available through an API. The closest open-weights option here is GLM 5.3.

Can GPT-5.5 read images?›

It accepts text and images, PDF files as input.

What is GPT-5.5?›

GPT-5.5 is a closed language model from OpenAI, released Apr 2026, the successor to GPT-5.4.

What are the best alternatives to GPT-5.5?›

Gemini 3.8 Flash (cheaper), GLM 5.3 (open weights).