OpenAI: o3

openai/o3

o3 (OpenAI) ranks #77 of 235 on the TryAii Score while still priced at $2/$8 per million input/output tokens — models at its level now sell for about $2.95 per million output tokens. DeepSeek V4 Flash 0423 matches it for $0.159 per million output tokens. It's strongest on math (#2 on MATH). It's among the fastest models we track, at roughly 66-108 tokens/sec.

Prompt $/1M

$2.000

Completion $/1M

$8.000

Context

200,000 tok

Best provider

129 tok/s · Artificial Analysis

Specs

Model id: openai/o3
Display name: OpenAI: o3
Modality: text+image+file->text
Tokenizer: GPT
Instruct type: —
Context length: 200,000
Max completion: 100,000
Prompt $/1M: $2.000
Completion $/1M: $8.000
Free tier: no

Capabilities

Function callingno
JSON modeno
System promptno
Visionno
Moderatedyes

Benchmark scores

Each axis is one benchmark, normalized to the best model in the field.

No benchmark data for this model.

Speed by provider

Provider	tok/s	TTFT	Total latency
Artificial Analysis	129	5,165 ms	—
OpenAIfastest	45	4,041 ms	4,831 ms
openaifastest	—	—	—

Prompt caching by provider

Provider	Min tokens to cache	Notes
OpenAI	1,024 tokens	Caching-eligible (gpt-4o-and-newer rule) but the cached-read discount is per-model (50%-98.75% platform range) and was not published for these models in the 2026-07 research — verify live pricing.