LLM Reference

Kimi K2.5 vs Qwen3.5-122B-A10B

Kimi K2.5 (2026) and Qwen3.5-122B-A10B (2026) compare a coding-specialized model against a standalone API model. Kimi K2.5 ships a 256k-token context window, while Qwen3.5-122B-A10B ships a 262k-token context window. On MMLU PRO, Kimi K2.5 leads by 0.4 pts. On pricing, Qwen3.5-122B-A10B costs $0.26/1M input tokens versus $0.44/1M for the alternative. This page treats the result as workflow and deployment fit, not a universal model winner.

Treat this as a product-type comparison: Kimi K2.5 is coding-specialized model, while Qwen3.5-122B-A10B is standalone API model. Choose based on workflow fit before reading any benchmark or price row as decisive.

Decision scorecard

Local evidence first
SignalKimi K2.5Qwen3.5-122B-A10B
Product typeCoding-specialized modelStandalone API model
Best forcustom coding agents, code generation, and tool loopsreasoning-heavy apps, multimodal apps, and tool-calling agents
Decision fitCoding, RAG, and AgentsCoding, RAG, and Agents
Context window256k262k
Cheapest output$2/1M tokens$2.08/1M tokens
Provider routes10 tracked3 tracked
Shared benchmarksMMLU PRO leader10 shared

Decision tradeoffs

Choose Kimi K2.5 when...
  • Kimi K2.5 holds a shared-benchmark lead on MMLU PRO, ahead by 0.4 points.
  • Kimi K2.5 has the lower cheapest tracked output price at $2/1M tokens.
  • Kimi K2.5 has broader tracked provider coverage for fallback and route flexibility.
  • Local decision data tags Kimi K2.5 for Coding, RAG, and Agents.
Choose Qwen3.5-122B-A10B when...
  • Qwen3.5-122B-A10B holds a shared-benchmark lead on Berkeley Function Calling Leaderboard v3, ahead by 7.7 points.
  • Qwen3.5-122B-A10B has the larger context window for long prompts, retrieval packs, or transcript analysis.
  • Qwen3.5-122B-A10B uniquely exposes Reasoning in local model data.
  • Local decision data tags Qwen3.5-122B-A10B for Coding, RAG, and Agents.

Monthly cost at traffic

Estimate token spend from the cheapest tracked input and output route or tier on this page.

Lower estimate Qwen3.5-122B-A10B

Kimi K2.5

$852

Cheapest tracked route/tier: OpenRouter

Qwen3.5-122B-A10B

$728

Cheapest tracked route/tier: OpenRouter

Estimated monthly gap: $124. Batch, cache, alternate speed tiers, and negotiated pricing are excluded from this local estimate.

Switch friction

Kimi K2.5 -> Qwen3.5-122B-A10B
  • Provider overlap exists on OpenRouter and Novita AI; start route-level A/B tests there.
  • Qwen3.5-122B-A10B is $0.08/1M tokens higher on cheapest tracked output pricing, so quality gains need to justify the spend.
  • Qwen3.5-122B-A10B adds Reasoning in local capability data.
Qwen3.5-122B-A10B -> Kimi K2.5
  • Provider overlap exists on OpenRouter and Novita AI; start route-level A/B tests there.
  • Kimi K2.5 is $0.08/1M tokens lower on cheapest tracked output pricing before cache, batch, or negotiated discounts.
  • Check replacement coverage for Reasoning before moving production traffic.

Specs

Specification
Released2026-03-152026-02-24
Context window256k262k
Parameters1T (MoE, 384 experts)122B
ArchitectureMixture of ExpertsMixture of Experts
LicenseProprietaryApache 2.0OSI-approved
OpennessProprietaryOpen source
WeightsNot releasedAvailable
CodeUnknownUnknown
Commercial useCommercial use: conditionalCommercial use: permitted
Knowledge cutoff--

Pricing and availability

Pricing attributeKimi K2.5Qwen3.5-122B-A10B
Input price$0.44/1M tokens$0.26/1M tokens
Output price$2/1M tokens$2.08/1M tokens
Providers

Capabilities

CapabilityKimi K2.5Qwen3.5-122B-A10B
VisionYesYes
MultimodalYesYes
ReasoningNoYes
JSON / Tool useYesYes
Structured outputsYesYes
Code executionNoNo
IDE integrationNoNo
Computer useNoNo
Parallel agentsNoNo

Benchmarks

BenchmarkKimi K2.5Qwen3.5-122B-A10B
MMLU PRO87.186.7
SWE-bench Verified76.872.0
Google-Proof Q&A87.985.7
LiveCodeBench85.078.9
Humanity's Last Exam50.225.3
Berkeley Function Calling Leaderboard v364.572.2
τ-bench74.279.5
BrowseComp60.663.8
MMMU Pro78.576.9
MultiChallenge61.461.5

Continue comparing

Last reviewed: 2026-06-29. Data sourced from public model cards and provider documentation.