o3 vs DeepSeek-R1

o3 is the better all-round choice, scoring 58/100 against DeepSeek-R1 (51). It leads on capability, inputs & features and context window. DeepSeek-R1 wins on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

DeepSeek-R1 is cheaper at $0.70 input / $2.60 output per million tokens (median across 11 API providers). o3 costs $2.00 input / $8.00 output per million tokens (official OpenAI API price). At a typical mix of three input tokens to one output token, that is $1.18 per million tokens for DeepSeek-R1 versus $3.50 for o3 (3× as much).

o3 scores higher on the Capabilities Index (ECI): o3 146.9 (#63 of 148) and DeepSeek-R1 139.0 (#104 of 148). Their confidence ranges do not overlap (144.9–148.6 vs 136.2–140.4), so the gap is a real one. On individual benchmarks: GPQA Diamond — o3 81.8%, DeepSeek-R1 71.7%; OTIS Mock AIME 2024–2025 — o3 84.4%, DeepSeek-R1 53.3%.

There are no published SWE-bench Verified results for DeepSeek-R1 yet, so there is no like-for-like coding score. On overall capability, o3 leads, which tends to carry over to coding, but test on your own codebase. Both support tool calling for agent workflows.

o3 has the largest context window at 200,000 tokens, against 128,000 for DeepSeek-R1. Maximum output per response: o3 up to 100,000, DeepSeek-R1 up to 32,768 tokens.

o3 accepts text, images and PDFs; DeepSeek-R1 accepts text. o3 handles the widest range of inputs.

DeepSeek-R1 publishes its weights and can be self-hosted; o3 is proprietary.

o3 is the newest, released Apr 16, 2025. DeepSeek-R1 came out Jan 20, 2025. Knowledge cutoff: o3 May 2024, DeepSeek-R1 Jul 2024.

Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Oct 4, 2026.