GLM-4.7 vs Kimi K2 Thinking vs Qwen3 14B

It is close. Our weighted score puts them within 2 points (Kimi K2 Thinking 58/100, GLM-4.7 56/100, Qwen3 14B 54/100), so choose by what matters most for your work: Kimi K2 Thinking for raw capability and Qwen3 14B on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

Qwen3 14B is cheaper at $0.35 input / $1.40 output per million tokens (official Alibaba API price). GLM-4.7 costs $0.60 input / $2.20 output per million tokens (official Z.AI API price); Kimi K2 Thinking costs $0.60 input / $2.50 output per million tokens (median across 10 API providers). At a typical mix of three input tokens to one output token, that is $0.613 per million tokens for Qwen3 14B versus $1.00 for GLM-4.7 (1.6× as much) and $1.07 for Kimi K2 Thinking (1.8× as much).

Kimi K2 Thinking scores higher on the Capabilities Index (ECI): Kimi K2 Thinking 146.0 (#72 of 148), GLM-4.7 143.5 (#84 of 148) and Qwen3 14B 138.2 (#107 of 148). The confidence ranges of the top two overlap (143.4–147.6 vs 141.3–145.6), so treat the gap as small. On individual benchmarks: GPQA Diamond — Kimi K2 Thinking 84.2%, GLM-4.7 83.3%, Qwen3 14B 63.8%; OTIS Mock AIME 2024–2025 — GLM-4.7 83.3%, Kimi K2 Thinking 83.1%, Qwen3 14B 66.4%.

There are no published SWE-bench Verified results for GLM-4.7, Kimi K2 Thinking and Qwen3 14B yet, so there is no like-for-like coding score. On overall capability, Kimi K2 Thinking leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.

Kimi K2 Thinking has the largest context window at 262,144 tokens, against 204,800 for GLM-4.7 and 131,072 for Qwen3 14B. Maximum output per response: GLM-4.7 up to 131,072, Kimi K2 Thinking up to 262,144, Qwen3 14B up to 8,192 tokens.

GLM-4.7 accepts text; Kimi K2 Thinking accepts text; Qwen3 14B accepts text. They handle the same number of input types.

Yes, all three publish their weights, so you can self-host them.

GLM-4.7 is the newest, released Dec 22, 2025. Kimi K2 Thinking came out Nov 6, 2025; Qwen3 14B came out Apr 29, 2025. Knowledge cutoff: GLM-4.7 Apr 2025, Kimi K2 Thinking Aug 2024, Qwen3 14B 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.