As of August 20, 2026, Kimi K2.5 is 33% cheaper per million tokens than Qwen3 Max. Kimi K2.5 is $0.59 / $3.09 per million input/output tokens versus Qwen3 Max at $0.50 / $5.00. Sources: https://platform.kimi.com/docs/pricing/chat and https://bailian.console.aliyun.com/
Sources: Moonshot AI pricing (https://platform.kimi.com/docs/pricing/chat); Qwen pricing (https://bailian.console.aliyun.com/)
Kimi K2.5
Moonshot AI
Input
$0.59
Output
$3.09
Qwen3 Max
Qwen
Input
$0.50
Output
$5.00
Token pricing, specs, and benchmarks side by side
Kimi K2.5Qwen3 MaxVerdict
Kimi K2.5 offers both lower pricing and stronger benchmark performance across the board, making it the clear value leader in this comparison.
As of August 20, 2026, Kimi K2.5 is 33% cheaper on blended LLM API pricing. Kimi K2.5 costs $0.59 per million input tokens and $3.09 per million output tokens ($3.68 blended 1M-in + 1M-out). Qwen3 Max costs $0.50 / $5.00 per million tokens ($5.50 blended). Sources: https://platform.kimi.com/docs/pricing/chat and https://bailian.console.aliyun.com/. The cheapest LLM for your app still depends on how many output tokens you generate.
Kimi K2.5 Moonshot AI API pricing is $0.59 input and $3.09 output per million tokens via Moonshot AI. Qwen3 Max Qwen API pricing is $0.50 input and $5.00 output per million tokens via Qwen. Use those four numbers, not ChatGPT Plus or Claude Pro subscription prices, when you are comparing APIs.
Kimi K2.5 leads on SWE-bench Verified: Kimi K2.5 at 76.8% vs Qwen3 Max at 69.6%. SWE-bench measures real GitHub fixes; LiveCodeBench measures contest programming.
Kimi K2.5 supports a 262K token context window, while Qwen3 Max supports 256K tokens. Kimi K2.5 offers a larger context window. Filling a larger window bills more input tokens, so the cheaper-per-million model can still cost more on long documents.
Kimi K2.5 leads on 6 of 6 shared benchmarks versus Qwen3 Max's 0 wins. Check the comparison table for GPQA Diamond, SWE-bench, MMLU, HLE, and the other evals we track. Token price and benchmark score together are the usual LLM comparison, not either number alone.
Both Kimi K2.5 and Qwen3 Max are production API models. For cost-sensitive production traffic, Kimi K2.5 has better economics on the blended 1M-in + 1M-out scale. For maximum capability, weight the benchmark table for your domain. Many production stacks route cheap models for drafts and a frontier model for the hard turn, which is usually cheaper than sending everything to the expensive API.
Yes. Call Kimi K2.5 through Moonshot AI and Qwen3 Max through Qwen with separate API keys, or through a gateway that already wraps both. Keep prompts in tokens, not characters, when you estimate the invoice. AnotherWrapper templates can swap providers without rewriting auth and billing.
List rates are the latest published Moonshot AI API pricing and Qwen API pricing, quoted per million tokens. Official rate sheets: https://platform.kimi.com/docs/pricing/chat and https://bailian.console.aliyun.com/. Prompt caching, batch APIs, and committed-use discounts can change the invoice. Benchmark scores are from official publications and independent evals, updated as new numbers land.
Kimi K2.5 supports up to 262K output tokens per request, while Qwen3 Max supports up to 131K output tokens. Output tokens are usually the expensive half of API pricing, so a higher max output cap is a capability, not a discount.
Neither Kimi K2.5 nor Qwen3 Max has published throughput data on this page yet. Throughput (tokens per second) and time-to-first-token are why two models with the same per-million-token price can still feel different in chat and agents.
Use the model pickers at the top of this LLM comparison. Search any model in the index and the URL updates. Open Kimi K2.5 or Qwen3 Max from the sidebar for the single-model API pricing page, or go back to the LLM pricing table to sort by cheapest blended cost.
This comparison covers API pricing, context window, throughput, and benchmark scores. Kimi K2.5 costs $0.59/M input and $3.09/M output. Qwen3 Max costs $0.50/M input and $5.00/M output.
Rank both on the LLM leaderboard, SWE-bench, and Humanity's Last Exam.
The index
280 models across 33 providers. Search or jump to a lab — every model page stays linked here.
Next step
Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.
