DeepSeek-R1 vs o3 vs Qwen3 235B-A22B

o3 is the better all-round choice, scoring 58/100 against Qwen3 235B-A22B (51) and DeepSeek-R1 (51). It leads on capability, inputs & features and context window. 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). Qwen3 235B-A22B costs $0.70 input / $2.80 output per million tokens (official Alibaba API price); 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 $1.23 for Qwen3 235B-A22B (1× as much) and $3.50 for o3 (3× as much).

o3 scores higher on the Capabilities Index (ECI): o3 146.9 (#63 of 148), Qwen3 235B-A22B 139.4 (#103 of 148) and DeepSeek-R1 139.0 (#104 of 148). Their confidence ranges do not overlap (144.9–148.6 vs 135.2–140.8), so the gap is a real one. On individual benchmarks: GPQA Diamond — o3 81.8%, DeepSeek-R1 71.7%, Qwen3 235B-A22B 70.7%.

There are no published SWE-bench Verified results for DeepSeek-R1 and Qwen3 235B-A22B 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. All three support tool calling for agent workflows.

o3 has the largest context window at 200,000 tokens, against 131,072 for Qwen3 235B-A22B and 128,000 for DeepSeek-R1. Maximum output per response: DeepSeek-R1 up to 32,768, o3 up to 100,000, Qwen3 235B-A22B up to 16,384 tokens.

DeepSeek-R1 accepts text; o3 accepts text, images and PDFs; Qwen3 235B-A22B accepts text. o3 handles the widest range of inputs.

DeepSeek-R1 and Qwen3 235B-A22B publishes its weights and can be self-hosted; o3 is proprietary.

Qwen3 235B-A22B is the newest, released Apr 28, 2025. o3 came out Apr 16, 2025; DeepSeek-R1 came out Jan 20, 2025. Knowledge cutoff: DeepSeek-R1 Jul 2024, o3 May 2024, Qwen3 235B-A22B Apr 2025.

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.