Long documents
Prompts that approach the documented context limit
Claude Opus 5.5
Claude Opus 5.5 has the larger documented context window.
Updated September 23, 2026. Rank says Claude Opus 5.5 is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Both of these models will change. Get the price, version and retirement notices for the pair, each with its source. Follow model changes
Claude Opus 5.5 has the higher public score, 88.45 versus 53.79, and the 90% score intervals do not overlap. 2 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Prompts that approach the documented context limit
Claude Opus 5.5
Claude Opus 5.5 has the larger documented context window.
1K fresh input + 500 output tokens
Kimi K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
200K cached + 20K fresh input + 10K output tokens
Kimi K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.
50K fresh input + 3K output tokens
Kimi K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Code generation, repair, and software-engineering tasks
Not enough matched evidence
Kimi K2.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Kimi K2.5 is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
Directional only · BenchAlign v5.6
Claude Opus 5.5 scores higher, but at least one score rests on Estimated evidence or a different benchmark set. Directional only, no winner.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
4 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
SWE-bench ProCoding
Normalized gap 39.2SWE MultilingualCoding
Normalized gap 20.9Each row shows the public-lane category score for both models: the BenchAlign v5.6 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
| Category | Claude Opus 5.5 | Kimi K2.5 | Basis | Reading |
|---|---|---|---|---|
| Agentic | 88.8Supported · #1/105 | 35.5Estimated · #54/105 | Directional onlyBenchAlign v5.6 lane · 11 vs 14 public rows | Directional only |
| Coding | 87.1Supported · #1/135 | 40.9Estimated · #58/135 | Directional onlyBenchAlign v5.6 lane · 9 vs 8 public rows | Directional only |
| Multimodal | 88.8#3/50 | 65.8#25/50 | Directional onlyProvisional lane · 1 vs 1 weighted rows | Directional only |
| Knowledge | 90.0Supported · #1/160 | 46.4Estimated · #68/160 | Directional onlyBenchAlign v5.6 lane · 16 vs 6 public rows | Directional only |
| Reasoning | 78.5#4/18 | 54.1Unranked · 3 rankable rows | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | 38.2#8/12 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Instruction following | Not ranked | 84.4#41/124 | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | Not ranked | 62.1#4/7 | Not comparableProvisional lane · 0 vs 4 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.6) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
1K fresh input + 500 output tokens
Kimi K2.5 has the lower modeled cost
Costs use the listed standard API rates.
50K fresh input + 3K output tokens
Kimi K2.5 has the lower modeled cost
Costs use the listed standard API rates.
200K cached + 20K fresh input + 10K output tokens
Kimi K2.5 has the lower modeled cost
Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
Maximum documented context; output-token limits may be lower.
Claude Opus 5.5
Kimi K2.5
256K
Claude Opus 5.5
claude-opus-5-5
Anthropic Claude Opus 5.5 model documentationKimi K2.5
Not sourced
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Opus 5.5
$0.2 per 1M cached input tokens
Claude Opus 5.5 model documentationKimi K2.5
Not published
Claude Opus 5.5
Kimi K2.5
Not sourced
Claude Opus 5.5
Kimi K2.5
Not sourced
Claude Opus 5.5
Generally Available · Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, Claude Platform on AWS
Anthropic Claude Opus 5.5 model documentationKimi K2.5
Not sourced
Claude Opus 5.5
Reasoning
Kimi K2.5
Non-Reasoning
Claude Opus 5.5
Proprietary
Kimi K2.5
Open Weight
Claude Opus 5.5
Proprietary
Kimi K2.5
Open Weight
Claude Opus 5.5
2026-09-22
Kimi K2.5
2026-02-01
Claude Opus 5.5 has the higher public score, 88.45 versus 53.79, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.
Claude Opus 5.5 scores higher for coding on the public lane, 87.1 to 40.9. Kimi K2.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Claude Opus 5.5 scores higher for agentic tasks on the public lane, 88.8 to 35.5. Kimi K2.5 is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
For the stated presets, chat costs $0.014 on Claude Opus 5.5 and $0.0021 on Kimi K2.5; repository review costs $0.26 and $0.039; the cache-heavy agent loop costs $0.32 and $0.162. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.
Claude Opus 5.5 has the larger documented context window: 1M, compared with 256K.
Estimates at 50,000 req/day · 1000 tokens/req average.
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Terminal-Bench 4.0
Not directly comparable
Terminal-Bench-Science 0.1
Not directly comparable
AutomationBench
Not directly comparable
HLE w/ tools
Not directly comparable
OSWorld 2.0
Not directly comparable
LAB all-pass (Harvey held-out)
Not directly comparable
LAB criterion-pass (Harvey held-out)
Not directly comparable
Toolathlon-Verified
Not directly comparable
Toolathlon Verified Pass@3
Not directly comparable
Toolathlon Verified Pass³
Not directly comparable
Toolathlon Verified avg. turns
Not directly comparable
Terminal-Bench 2.0
Not directly comparable
BrowseComp
Not directly comparable
Claw-Eval
Not directly comparable
QwenClawBench
Not directly comparable
τ³-bench results
Not directly comparable
DeepSearchQA
Not directly comparable
DeepPlanning
Not directly comparable
Toolathlon
Not directly comparable
MCP Atlas
Not directly comparable
MCP-Tasks
Not directly comparable
WideResearch
Not directly comparable
Gert Labs
Not directly comparable
ResearchClawBench
Not directly comparable
JobBench
Not directly comparable
FrontierCode 1.1 Main
Not directly comparable
cursorBench40
Not directly comparable
SWE-bench Pro
Claude Opus 5.5 leads this result
SWE Multilingual
Claude Opus 5.5 leads this result
SWE Multimodal
Not directly comparable
DeepSWE
Not directly comparable
FrontierCode 1.1 Extended
Not directly comparable
FrontierSWE v2
Not directly comparable
ProgramBench
Not directly comparable
SWE-bench Verified
Not directly comparable
SWE-bench Verified*
Not directly comparable
LiveCodeBench v6
Not directly comparable
SWE-Rebench
Not directly comparable
React Native Evals
Not directly comparable
SciCode
Not directly comparable
LongBench v2
Not directly comparable
Chartography (tools)
Not directly comparable
Chartography (no tools)
Not directly comparable
BenchCAD Vision2Code (no tools)
Not directly comparable
BenchCAD Vision2Code (tools)
Not directly comparable
Biomedical image analysis
Not directly comparable
OfficeQA
Not directly comparable
OfficeQA Pro
Not directly comparable
MMMU-Pro
Not directly comparable
Video-MME
Not directly comparable
MMVU
Not directly comparable
VideoMMMU
Not directly comparable
HLE w/o tools
Not directly comparable
HealthBench (raw)
Not directly comparable
HealthBench (length-adjusted)
Not directly comparable
HealthBench Professional
Not directly comparable
HealthBench Professional (raw)
Not directly comparable
BioMysteryBench (human-solvable)
Not directly comparable
BioMysteryBench (human-difficult)
Not directly comparable
SpatialBench Verified
Not directly comparable
SingleCellBench
Not directly comparable
Morphology-to-molecule matching
Not directly comparable
Medicinal chemistry
Not directly comparable
Protein Design
Not directly comparable
Protein Design library ranking
Not directly comparable
De novo protein-binder design
Not directly comparable
Protocols (troubleshooting)
Not directly comparable
Protocols (understanding)
Not directly comparable
GPQA
Not directly comparable
GPQA-D
Not directly comparable
SuperGPQA
Not directly comparable
MMLU-Pro
Not directly comparable
MMLU-Pro (Arcee)
Not directly comparable
HLE
Not directly comparable
IFEval
Not directly comparable
ArXivMath Aug. 2026 (no tools)
Not directly comparable
ArXivMath Aug. 2026 (tools)
Not directly comparable
AIME 2025
Not directly comparable
AIME26
Not directly comparable
AIME25 (Arcee)
Not directly comparable
HMMT Feb 2025
Not directly comparable
HMMT Nov 2025
Not directly comparable
HMMT Feb 2026
Not directly comparable
MMAnswerBench
Not directly comparable
FrontierMath v2 (Tiers 1-3)
Not directly comparable
FrontierMath v2 (Tier 4)
Not directly comparable
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.
Last updated September 23, 2026