Updated October 13, 2026. Rank says Hy4 preview 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
Hy4 preview has the higher public point estimate, 60.86 versus 47.89. Their conditional score ranges overlap. These ranges do not establish rank confidence. 6 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.
Code generation, repair, and software-engineering tasks
Hy4 preview
Hy4 preview has the higher public coding point estimate, 57.1 to 34.9, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Prompts that approach the documented context limit
Hy4 preview
Hy4 preview has the larger documented context window.
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Hy4 preview is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. MiniMax M2.7 does not fit this workload in one request. MiniMax M2.7 has no published cached-input rate, so cached tokens use its listed input rate. Hy4 preview has no comparable published API token rate.
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
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 v5.8
Hy4 preview 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.
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
2 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.
HLEKnowledge
Normalized gap 27.4SWE-bench ProCoding
Normalized gap 9.5SWE MultilingualCoding
Normalized gap 6.4Each row shows the public-lane category score for both models: the BenchAlign v5.8 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 | Hy4 preview | MiniMax M2.7 | Basis | Reading |
|---|---|---|---|---|
| Coding | 57.1Supported · #26/148 | 34.9Supported · #78/148 | Like-for-likeBenchAlign v5.8 lane · 8 vs 12 public rows | Hy4 preview leads |
| Agentic | 54.5Estimated · #37/123 | 31.4Supported · #86/123 | Directional onlyBenchAlign v5.8 lane · 15 vs 8 public rows | Directional only |
| Knowledge | 63.6Estimated · #39/180 | 45.7Supported · #92/180 | Directional onlyBenchAlign v5.8 lane · 4 vs 5 public rows | Directional only |
| Reasoning | Not ranked | 79.6Unranked · 2 rankable rows | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multimodal | 88.3Unranked · 1 rankable row | Not ranked | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Instruction following | Not ranked | 93.0#13/127 | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | Not ranked | 70.5Unranked · 4 rankable rows | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) 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
Hy4 preview has no comparable published API token rate.
50K fresh input + 3K output tokens
Hy4 preview has no comparable published API token rate.
200K cached + 20K fresh input + 10K output tokens
MiniMax M2.7 does not fit this workload in one request. MiniMax M2.7 has no published cached-input rate, so cached tokens use its listed input rate. Hy4 preview has no comparable published API token 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.
Hy4 preview
MiniMax M2.7
200K
Hy4 preview
tencent/Hy4-preview
Tencent Hy4 preview model cardMiniMax M2.7
Not sourced
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Hy4 preview
No comparable hosted API rate
Tencent Hy4 preview model cardMiniMax M2.7
Not published
Hy4 preview
Not sourced
MiniMax M2.7
Not sourced
Hy4 preview
Not sourced
MiniMax M2.7
Not sourced
Hy4 preview
Not sourced
MiniMax M2.7
Not sourced
Hy4 preview
Reasoning
MiniMax M2.7
Non-Reasoning
Hy4 preview
Open Weight
MiniMax M2.7
Open Weight
Hy4 preview
Open Weight
MiniMax M2.7
Open Weight
Hy4 preview
2026-08-28
MiniMax M2.7
2026-03-18
Hy4 preview has the higher public point estimate, 60.86 versus 47.89. Their conditional score ranges overlap. These ranges do not establish rank confidence. The page therefore keeps the decision tied to the specific documented workload.
Hy4 preview has the higher public coding point estimate, 57.1 to 34.9, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Hy4 preview scores higher for agentic tasks on the public lane, 54.5 to 31.4. Hy4 preview 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.
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Hy4 preview has the larger documented context window: 1M, compared with 200K.
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Terminal-Bench 2.1
Not directly comparable
CyberGym
Not directly comparable
WideResearch
Not directly comparable
DRACO
Not directly comparable
MCP Atlas
Not directly comparable
Toolathlon-Verified
Not directly comparable
APEX-Agents
Not directly comparable
skillsBench
Not directly comparable
JobBench
Not directly comparable
Agents' Last Exam
Not directly comparable
AutomationBench
Not directly comparable
BankerToolBench
Not directly comparable
HLE w/ tools
Not directly comparable
Terminal-Bench 2.1 (Vals)
Hy4 preview leads this result
CWE-bench v1
Not directly comparable
Terminal-Bench 2.0
Not directly comparable
τ³-bench results
Not directly comparable
Toolathlon
Not directly comparable
MLE-Bench Lite
Not directly comparable
MM-ClawBench
Not directly comparable
Claw-Eval
Not directly comparable
Gert Labs
Not directly comparable
Terminal-Bench 2.1
Not directly comparable
SWE-bench Pro
Hy4 preview leads this result
SWE Multilingual
Hy4 preview leads this result
DeepSWE
Not directly comparable
NL2Repo
Hy4 preview leads this result
ProgramBench
Not directly comparable
PostTrain Bench
Not directly comparable
sweMarathon
Not directly comparable
SWE-bench Verified*
Not directly comparable
SWE-Rebench
Not directly comparable
Multi-SWE Bench
Not directly comparable
VIBE-Pro
Not directly comparable
Terminal-Bench 2.0
Not directly comparable
Vibe Code Bench
Not directly comparable
React Native Evals
Not directly comparable
LiveCodeBench (Vals)
Not directly comparable
SWE-bench (Vals)
Not directly comparable
CritPt
Not directly comparable
OfficeQA Pro
Not directly comparable
GPQA
Not directly comparable
GPQA-D
Hy4 preview leads this result
HLE
Hy4 preview leads this result
HLE w/o tools
Not directly comparable
MMLU-Pro (Arcee)
Not directly comparable
GPQA Diamond (Vals)
Not directly comparable
MMLU-Pro (Vals)
Not directly comparable
Apex
Not directly comparable
AIME25 (Arcee)
Not directly comparable
AIME26
Not directly comparable
HMMT Nov 2025
Not directly comparable
HMMT Feb 2026
Not directly comparable
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 5,500+ readers.
Last updated October 13, 2026
