vs
8:1
overall capability wins (0 ties across 9 scored). LongCat-2.0 leads overall.
- Reasoning5:0
- Coding1:1
- Tool use1:0
- Composite indices2:0
- Specs1:0
Verdict
LongCat-2.0 edges reasoning & knowledge; LongCat-2.0 offers ~7.8x larger context (1M vs 128K); LongCat-2.0 ships open weights (MIT).
- LongCat-2.0 edges reasoning & knowledge
- LongCat-2.0 offers ~7.8x larger context (1M vs 128K)
- LongCat-2.0 ships open weights (MIT)
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- LongCat-2.0
- -
- Mercury 2
- $0.44
Price per 1M tokens
Full comparison
Organization
Meituan
Inception Labs
Family
Other
Other
License
MIT
Proprietary
Open weights
Yes
No
Release
Jun 29, 2026
Sep 1, 2026
Knowledge cutoff
-
-
API / provider
No primary API price listed
Inception Labs
Modalities
text → text
text → text
Specs
Context window
1M
128K
Max output
64K
-
Parameters
1.6T (48B act.)
—
Pricing
Input $/1M
—
$0.25
Output $/1M
—
$0.75
Blended $/1M (3∶1)
-
$0.38
Speed
tok/s
55
-
TTFT (s)
0.5
-
Reasoning
MMLU-Pro
—
—
GPQA Diamond
78.0
77.0
Humanity's Last Exam
33.7
17.1
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
29.6
21.2
AA-LCR v1.1
65.0
43.7
CritPt
2.6
0.8
MMMU-Pro
—
—
IFBench
—
69.8
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
50.2
27.3
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
36.3
37.7
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
1,167
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
—
| Metric | LongCat-2.0 | Mercury 2 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Meituan | Inception Labs | — |
| Family | Other | Other | — |
| License | MIT | Proprietary | — |
| Open weights | Yes | No | — |
| Release | Jun 29, 2026 | Sep 1, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | No primary API price listed | Inception Labs | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 1M | 128K | A +872K |
| Max output | 64K | - | — |
| Parameters | 1.6T (48B act.) | — | — |
| Pricing | |||
| Input $/1M | — | $0.25 | — |
| Output $/1M | — | $0.75 | — |
| Blended $/1M (3∶1) | - | $0.38 | — |
| Speed | |||
| tok/s | 55 | - | — |
| TTFT (s) | 0.5 | - | — |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 78.0 | 77.0 | A +1.0 pts |
| Humanity's Last Exam | 33.7 | 17.1 | A +16.6 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 29.6 | 21.2 | A +8.4 pts |
| AA-LCR v1.1 | 65.0 | 43.7 | A +21.3 pts |
| CritPt | 2.6 | 0.8 | A +1.8 pts |
| MMMU-Pro | — | — | — |
| IFBench | — | 69.8 | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 50.2 | 27.3 | A +22.9 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 36.3 | 37.7 | B +1.4 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | 1,167 | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | — | — |
LongCat-2.0
- Context
- 1M / 64K out
- Parameters
- 1.6T (48B act.)
- Price
- —
- Speed
- 55 tok/s · 0.5s TTFT
- Modalities
- text → text
- License
- MIT
1.6T MoE trained on domestic Chinese chips - open MIT weights focused on agentic coding.
Mercury 2
- Context
- 128K
- Parameters
- —
- Price
- $0.25 / $0.75
- Speed
- —
- Modalities
- text → text
- License
- Proprietary
Inception Labs' diffusion-based (non-autoregressive) reasoning LLM, API-only with a 128K context window and high-throughput inference on Blackwell GPUs.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.