Coding work
Code generation, repair, and software-engineering tasks
Qwen4 Max
Qwen4 Max has the higher public coding point estimate, 76.1 to 67.9, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Updated November 21, 2026. Rank says Qwen4 Max is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Qwen4 Max has the higher public point estimate, 74.8 versus 63.5. Their conditional score ranges overlap. These ranges do not establish rank confidence. 3 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Claude Haiku 5.5: $0.10 input / $0.50 output
Qwen4 Max: $1.50 input / $4.50 output
Claude Haiku 5.5: 1M
Qwen4 Max: 1M
Claude Haiku 5.5: October 7, 2026
Qwen4 Max: November 16, 2026
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.
Code generation, repair, and software-engineering tasks
Qwen4 Max
Qwen4 Max has the higher public coding point estimate, 76.1 to 67.9, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Tool use, computer use, and multi-step task completion
Qwen4 Max
Qwen4 Max has the higher public agentic point estimate, 75.1 to 68.3, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
1K fresh input + 500 output tokens
Claude Haiku 5.5
Claude Haiku 5.5 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
Claude Haiku 5.5
Claude Haiku 5.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
50K fresh input + 3K output tokens
Claude Haiku 5.5
Claude Haiku 5.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
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.
Like-for-like · BenchAlign v6.0
Qwen4 Max has the higher coding point estimate. Conditional score ranges do not establish rank confidence.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
The LLM Selector reads what you need a model for and returns a shortlist, ranked in the same order the job pages publish.
We keep what you type here, drafts included, for 180 days, and after that only the job and requirements read from it. Nothing is tied to your IP address or account. Leave out names and other personal details. Privacy
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.
HLE w/o toolsKnowledge
Normalized gap 5.8Each row shows the public-lane category score for both models: the BenchAlign v6.0 lane in every category, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence. Directional and not-comparable rows remain visible, but they never receive a winner.
| Category | Claude Haiku 5.5 | Qwen4 Max | Basis | Reading |
|---|---|---|---|---|
| Agentic | 68.3Supported · #25/182 | 75.1Supported · #17/182 | Like-for-likeBenchAlign v6.0 lane · 9 vs 1 public rows | Qwen4 Max leads · intervals overlap |
| Coding | 67.9Supported · #28/180 | 76.1Supported · #19/180 | Like-for-likeBenchAlign v6.0 lane · 7 vs 3 public rows | Qwen4 Max leads · intervals overlap |
| Reasoning | 62.6Supported · #38/161 | 75.7Estimated · #12/161 | Directional onlyBenchAlign v6.0 lane · 4 vs 0 public rows | Directional only |
| Multimodal | 63.9Supported · #39/119 | 73.4Estimated · #17/119 | Directional onlyBenchAlign v6.0 lane · 4 vs 1 public rows | Directional only |
| Knowledge | 50.2Supported · #84/158 | 67.9Estimated · #23/158 | Directional onlyBenchAlign v6.0 lane · 1 vs 3 public rows | Directional only |
| Instruction following | 58.3Supported · #37/143 | 72.5Estimated · #15/143 | Directional onlyBenchAlign v6.0 lane · 0 vs 0 public rows | Directional only |
| Multilingual | 56.1Estimated · #36/88 | Not ranked | Not comparableBenchAlign v6.0 lane · 0 vs 0 public rows | Not comparable |
| Math | 62.1Supported · #29/125 | Not ranked | Not comparableBenchAlign v6.0 lane · 2 vs 0 public rows | Not comparable |
Ranks count the models scored in each category’s BenchAlign v6.0 lane, so denominators differ by category. 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
Claude Haiku 5.5 has the lower modeled cost
Costs use the listed standard API rates.
50K fresh input + 3K output tokens
Claude Haiku 5.5 has the lower modeled cost
Costs use the listed standard API rates.
200K cached + 20K fresh input + 10K output tokens
Claude Haiku 5.5 has the lower modeled cost
Costs use the listed standard API rates.
Enter a month of tokens to price it at both models’ listed API rates.
Standard listed rates, no batch or volume discounts.
Overall score against output price per 1M tokens, one model per family. Pricing as of November 21, 2026.
Claude Haiku 5.5Qwen4 Max
Gray dots are the other priced models. The dark step line is the efficiency frontier: for a model under it, another model costs no more and scores at least as high.
See Claude Haiku 5.5 on the full price vs performance chartSourced 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 Haiku 5.5
Qwen4 Max
Claude Haiku 5.5
claude-haiku-5-5
Anthropic Claude Haiku 5.5 model documentationQwen4 Max
qwen4-max
Qwen Cloud API pricingA missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Haiku 5.5
$0.01 per 1M cached input tokens
Claude API pricingQwen4 Max
$0.075 per 1M cached input tokens
Qwen Cloud API pricingClaude Haiku 5.5
Qwen4 Max
Not sourced
Claude Haiku 5.5
Qwen4 Max
Not sourced
Claude Haiku 5.5
Generally Available · Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, Claude Platform on AWS
Anthropic Claude Haiku 5.5 model documentationQwen4 Max
Not sourced
Claude Haiku 5.5
Reasoning
Qwen4 Max
Reasoning
Claude Haiku 5.5
Proprietary
Qwen4 Max
Proprietary
Claude Haiku 5.5
Proprietary
Qwen4 Max
Proprietary
Claude Haiku 5.5
2026-10-07
Qwen4 Max
2026-11-16
Qwen4 Max has the higher public point estimate, 74.8 versus 63.5. Their conditional score ranges overlap. These ranges do not establish rank confidence. The page therefore keeps the decision tied to the specific documented workload.
Qwen4 Max has the higher public coding point estimate, 76.1 to 67.9, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Qwen4 Max has the higher public agentic tasks point estimate, 75.1 to 68.3, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
For the stated presets, chat costs $0.00035 on Claude Haiku 5.5 and $0.00375 on Qwen4 Max; repository review costs $0.0065 and $0.0885; the cache-heavy agent loop costs $0.009 and $0.09. Costs use the listed standard API rates.
Both models list the same context window, 1M.
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Gert Labs
Shared sourceQwen4 Max leads this result
Terminal-Bench 4.1
Not directly comparable
Agents' Last Exam (Linux)
Not directly comparable
Terminal-Bench 4.0
Not directly comparable
HLE w/ tools
Not directly comparable
MazeBench
Not directly comparable
Hermes Index
Not directly comparable
Hermes Bench
Not directly comparable
OSWorld 2.1 Offline
Not directly comparable
FrontierCode 1.1 Main
Not directly comparable
Bug Hunt Bench
Not directly comparable
FrontierSWE v2
Not directly comparable
FrontierCode 2.0
Not directly comparable
CursorBench 4.0
Not directly comparable
VulcanBench Frontier v4
Not directly comparable
CursorBench 4.1
Not directly comparable
DeepSWE
Not directly comparable
ProgramBench
Not directly comparable
PostTrainBench
Not directly comparable
ParseBench-2
Not directly comparable
PerceptionBench (tools)
Not directly comparable
ParseBench-1
Not directly comparable
Chartography (no tools)
Qwen4 Max leads this result
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 7,000+ subscribers.
Last updated November 21, 2026