Agentic
Like-for-like- Inkling-Small
- 36.9
- Supported · #128/152
- MiniMax M3
- 42.1
- Supported · #107/152
- Basis
- BenchAlign lane · 5 vs 9 public rows
- Reading
- MiniMax M3 leads · intervals overlap
Keep up with the models you depend on. Follow price changes, retirements, and API updates.Follow the models you depend on.
Follow model changesUpdated September 10, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
MiniMax M3 has the higher public score estimate, 61.55 versus 59.25, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
12 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Share or export
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.
Tool use, computer use, and multi-step task completion
MiniMax M3
MiniMax M3 leads on the public agentic lane, 42.1 to 36.9, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
1K fresh input + 500 output tokens
MiniMax M3
MiniMax M3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
200K cached + 20K fresh input + 10K output tokens
MiniMax M3
MiniMax M3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
50K fresh input + 3K output tokens
MiniMax M3
MiniMax M3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Code generation, repair, and software-engineering tasks
Not enough matched evidence
MiniMax M3 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documented
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
3 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Each row shows the public-lane category score for both models: the BenchAlign 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 | Inkling-Small | MiniMax M3 | Basis | Reading |
|---|---|---|---|---|
| Agentic | 36.9Supported · #128/152 | 42.1Supported · #107/152 | Like-for-likeBenchAlign lane · 5 vs 9 public rows | MiniMax M3 leads · intervals overlap |
| Knowledge | 56.5Supported · #46/183 | 53.2Supported · #64/183 | Like-for-likeBenchAlign lane · 6 vs 2 public rows | Inkling-Small leads · intervals overlap |
| Coding | 43.8Supported · #97/151 | 48.8Estimated · #66/151 | Directional onlyBenchAlign lane · 6 vs 10 public rows | Directional only |
| Multimodal | 48.8#39/48 | 52.2#34/48 | Directional onlyProvisional lane · 2 vs 2 weighted rows | Directional only |
| Instruction following | 89.6#25/123 | 93.7#4/123 | Directional onlyProvisional lane · 1 vs 0 weighted rows | Directional only |
| Reasoning | 42.7Unranked · 3 rankable rows | 78.0#4/20 | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Math | 76.9Unranked · 2 rankable rows | Not ranked | Not comparableProvisional lane · 2 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
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.
BrowseComp
Agentic
MMMU-Pro
Multimodal
LiveCodeBench (Vals)
Coding
SWE-bench Pro
Coding
MMLU-Pro (Vals)
Knowledge
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
MiniMax M3 has the lower modeled cost
Costs use the listed standard API rates.
50K fresh input + 3K output tokens
MiniMax M3 has the lower modeled cost
Costs use the listed standard API rates.
200K cached + 20K fresh input + 10K output tokens
MiniMax M3 has the lower modeled cost
Costs use the listed standard API rates.
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.
Inkling-Small
1M
MiniMax M3
1M
Inkling-Small
Not sourced
MiniMax M3
Not sourced
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Inkling-Small
$0.116 per 1M cached input tokens
MiniMax M3
$0.06 per 1M cached input tokens
Inkling-Small
Not sourced
MiniMax M3
Not sourced
Inkling-Small
Not sourced
MiniMax M3
Not sourced
Inkling-Small
Not sourced
MiniMax M3
Not sourced
Inkling-Small
Hybrid
MiniMax M3
Non-Reasoning
Inkling-Small
Open Weight
MiniMax M3
Open Weight
Inkling-Small
Open Weight
MiniMax M3
Open Weight
Inkling-Small
2026-07-30
MiniMax M3
2026-06-01
Run the same representative tasks against both endpoints before changing production traffic.
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Terminal-Bench 2.0
MiniMax M3 leads this result
BrowseComp
MiniMax M3 leads this result
MCP Atlas
Inkling-Small leads this result
Toolathlon-Verified
Not directly comparable
Terminal-Bench 2.1 (Vals)
Inkling-Small leads this result
OSWorld-Verified
Not directly comparable
Claw-Eval
Not directly comparable
BankerToolBench
Not directly comparable
ResearchClawBench
Not directly comparable
OSWorld 2.0
Not directly comparable
SWE-bench Verified
MiniMax M3 leads this result
SWE-bench Pro
MiniMax M3 leads this result
Terminal-Bench 2.0
MiniMax M3 leads this result
SciCode
Not directly comparable
LiveCodeBench (Vals)
Inkling-Small leads this result
SWE-bench (Vals)
Inkling-Small leads this result
NL2Repo
Not directly comparable
VIBE V2
Not directly comparable
SVG-Bench
Not directly comparable
KernelBench Hard
Not directly comparable
OpenHarmony Bench
Not directly comparable
GPQA
Not directly comparable
GPQA-D
Not directly comparable
HLE
Not directly comparable
HLE w/o tools
Not directly comparable
GPQA Diamond (Vals)
MiniMax M3 leads this result
MMLU-Pro (Vals)
Inkling-Small leads this result
MMMU-Pro
MiniMax M3 leads this result
CharXiv
Not directly comparable
CharXiv w/o tools
Not directly comparable
OfficeQA Pro
Not directly comparable
OmniDocBench 1.5
Not directly comparable
VideoMMMU
Not directly comparable
Video-MME (with subtitle)
Not directly comparable
IFBench
Not directly comparable
MiniMax M3 has the higher public score estimate, 61.55 versus 59.25, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
MiniMax M3 scores higher for coding on the public lane, 48.8 to 43.8. MiniMax M3 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.
MiniMax M3 leads the public agentic tasks lane, 42.1 to 36.9, with Supported evidence for both models, although the 90% intervals overlap.
For the stated presets, chat costs $0.0013 on Inkling-Small and $0.0009 on MiniMax M3; repository review costs $0.03332 and $0.0186; the cache-heavy agent loop costs $0.0492 and $0.03. Costs use the listed standard API rates.
Both models list the same context window, 1M.
Last updated September 10, 2026
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.