the ranking
| # | Model | SWE-bench Verified | Best for |
|---|---|---|---|
| 1 | DeepSeek V4 Pro openDeepSeek | 80.6% | The cheapest frontier-class coder — top open-weights score at ~11× less than Opus. Best pick when cost or self-hosting rules. |
| 2 | MiniMax M3 openMiniMax | 80.5% | Open weights with 1M context, multimodal input and computer use — beats GPT-5.5 on SWE-bench Pro at 5–10% of the cost. |
| 3 | Kimi K2.6 openMoonshot AI | 80.2% | A top-three open coder whose 58.6% SWE-bench Pro beats several closed flagships. |
the top picks, decoded
1. DeepSeek V4 Pro — 80.6% SWE-bench Verified
The cheapest frontier-class coder — top open-weights score at ~11× less than Opus. Best pick when cost or self-hosting rules. Vendor-reported, as aggregated by llm-stats (June 2026). Corrected Jul 17, 2026: we previously called llm-stats an "independent tracker", but llm-stats labels its own SWE-bench Verified table "Verified: 0 / Self-reported: 104" — every score on it is vendor-claimed, so this is a vendor number, not an independent one. No independent evaluator has run V4 Pro; vals.ai has evaluated the plain DeepSeek V4 (77.4%), which is a different model. Tied with Gemini 3.1 Pro on Verified, ahead on Pro.
2. MiniMax M3 — 80.5% SWE-bench Verified
Open weights with 1M context, multimodal input and computer use — beats GPT-5.5 on SWE-bench Pro at 5–10% of the cost. Vendor-reported at launch (Jun 1, 2026). Independent comparison: vals.ai's bash-only harness measures MiniMax-M3 at 75.0% ±1.94, 5.5 points lower, so treat the 80.5% as a best-case number.
3. Kimi K2.6 — 80.2% SWE-bench Verified
A top-three open coder whose 58.6% SWE-bench Pro beats several closed flagships. Vendor-reported (10-run average on Moonshot's own SWE-agent harness). Independent comparison: vals.ai's bash-only harness measures Kimi K2.6 at 76.2% ±1.91, 4 points lower. Price: official Moonshot API $0.95/$4 per 1M (cache-hit input $0.16).
how we rank
We rank by SWE-bench Verified (500 real, human-validated GitHub issues resolved end-to-end), tiebroken by the harder SWE-bench Pro. A score is only printed once confirmed against an independent evaluation or the maker's primary source — and every row states which kind it is. Where both exist, we print both: one as the ranked score, the other in that row's note. We would rather show you the gap than ask you to trust our pick. Our independent reference is vals.ai, which runs every model itself through the same minimal bash-only harness (mini-swe-agent), so the models are compared on equal footing. That matters more than it sounds: SWE-bench scores a model and its scaffolding together, and vendors report using their own tuned scaffolds. Against vals.ai's neutral harness, the vendor claims on this board run 2.6 to 11.6 points optimistic. So rows marked vendor-reported are best-case numbers and are not strictly comparable to the independent ones — where we know the independent figure, we print it in the row's note. That is a deliberate choice and you should know we made it: on four rows (Claude Sonnet 5, MiniMax M3, Qwen3.7 Max, Kimi K2.6) an independent score for that exact model exists and is lower, and we still rank on the maker's published figure because it is the number that model is sold and quoted on. We disclose the independent one beside it instead of quietly restating the board on a single evaluator's harness choice. The honest consequence: positions that straddle the two regimes are approximate. Qwen3.7 Max is the sharpest case — it sits at #11 on Alibaba's 80.4%, and on the neutral harness its 68.8% would put it far down the table. Note that llm-stats, which we previously miscredited as an independent tracker, labels its own SWE-bench Verified table "Verified: 0 / Self-reported: 104"; it aggregates vendor claims. Models still being checked are marked “verifying” and shown without a number rather than estimated. Prices are per 1M input tokens on the standard API tier and can change — always confirm current pricing with the provider.
Want the raw numbers? The full dataset is public: JSON · CSV.
From our full AI Coding Leaderboard (2026-07-17). We only rank scores confirmed against primary sources.