BenchLM recommendation
Best Value LLM for Math in 2026 — Cost-Adjusted Rankings
As of July 23, 2026, the top model in best value llm for math on the BenchLM leaderboard is Kimi K2.5 with a score of 20.9.
Last verified: July 23, 2026
Math is the most saturated benchmark category — top models all score 95%+ on competition math. That makes price the main differentiator for math-heavy workloads. This ranking divides each model's weighted math score by output token price. If you need strong math reasoning (AIME, BRUMO, MATH-500) and the top 10 models all deliver similar accuracy, the value ranking here helps you pick the most cost-effective option.
Unless noted otherwise, ranking surfaces on this page use BenchLM's provisional leaderboard lane rather than the stricter sourced-only verified leaderboard.
Bottom line: Math is the most saturated category — top models all score 95%+. That makes price the main differentiator. Gemini 3.1 Flash-Lite and DeepSeek Coder 2.0 lead on value.
Kimi K2.5 leads this ranking with a score of 20.93, followed by Kimi K2.6 (18.08) and GLM-5 (17.88). There is meaningful separation between the top models, suggesting genuine performance differences.
The best open-weight option is Kimi K2.5 (ranked #1 with a score of 20.93). Open-weight models are highly competitive in this category — self-hosting is a viable alternative to proprietary APIs.
This ranking is based on provisional weighted averages across the scoring benchmarks in math tracked by BenchLM.ai. For detailed model profiles, click any model name below. To compare two specific models head-to-head, use the "vs #" links.
What changed
Gemini 3.1 Flash-Lite leads math value — decent math scores at the lowest price.
DeepSeek Coder 2.0 strong raw math (71) at very low cost.
Gemini 2.5 Flash good math value with broader capabilities.
How to choose
Full Rankings (6 models)
Key Takeaways
The best value model is Kimi K2.5 by Moonshot AI with a provisional Score/$ ratio of 20.93 (score: 62.8, output: $3/1M tokens).
The best open-weight model is Kimi K2.5 at position #1.
6 models are included in this ranking.
Score in Context
What these scores mean
Value scores divide the weighted math score by output token price (per 1M tokens). Higher means more capability per dollar. Models with no listed price are excluded.
Known limitations
Value rankings favor cheap models even if absolute performance is modest. A model scoring half as well at one-tenth the price wins on value — but may not meet your quality bar. Always check raw scores alongside value rankings.
Explore More
Choose a model with this week’s evidence
Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.
One email each week. Unsubscribe anytime.