vs
2:0
overall capability wins (0 ties across 2 scored). Atria Dawn Preview leads overall.
- Tool use1:0
- Specs1:0
Verdict
Atria Dawn Preview offers ~2.0x larger context (256K vs 128K); Atria Dawn Preview ships open weights (MIT).
- Atria Dawn Preview offers ~2.0x larger context (256K vs 128K)
- Atria Dawn Preview ships open weights (MIT)
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Atria Dawn Preview
- -
- Mercury 2
- $0.44
Price per 1M tokens
Full comparison
Organization
Shanghai AI Lab
Inception Labs
Family
Other
Other
License
MIT
Proprietary
Open weights
Yes
No
Release
Sep 11, 2026
Sep 1, 2026
Knowledge cutoff
-
-
API / provider
No primary API price listed
Inception Labs
Modalities
text → text
text → text
Specs
Context window
256K
128K
Max output
66K
-
Parameters
744B
—
Pricing
Input $/1M
—
$0.25
Output $/1M
—
$0.75
Blended $/1M (3∶1)
-
$0.38
Speed
tok/s
-
-
TTFT (s)
-
-
Reasoning
MMLU-Pro
—
—
GPQA Diamond
—
77.0
Humanity's Last Exam
—
17.1
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
—
21.2
AA-LCR v1.1
—
43.7
CritPt
—
0.8
MMMU-Pro
—
—
IFBench
—
69.8
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
59.6
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
—
27.3
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
—
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 | Atria Dawn Preview | Mercury 2 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Shanghai AI Lab | Inception Labs | — |
| Family | Other | Other | — |
| License | MIT | Proprietary | — |
| Open weights | Yes | No | — |
| Release | Sep 11, 2026 | Sep 1, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | No primary API price listed | Inception Labs | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 256K | 128K | A +128K |
| Max output | 66K | - | — |
| Parameters | 744B | — | — |
| Pricing | |||
| Input $/1M | — | $0.25 | — |
| Output $/1M | — | $0.75 | — |
| Blended $/1M (3∶1) | - | $0.38 | — |
| Speed | |||
| tok/s | - | - | — |
| TTFT (s) | - | - | — |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | — | 77.0 | — |
| Humanity's Last Exam | — | 17.1 | — |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | — | 21.2 | — |
| AA-LCR v1.1 | — | 43.7 | — |
| CritPt | — | 0.8 | — |
| MMMU-Pro | — | — | — |
| IFBench | — | 69.8 | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | 59.6 | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | — | 27.3 | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | — | 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 | — | — | — |
Atria Dawn Preview
- Context
- 256K / 66K out
- Parameters
- 744B
- Price
- —
- Speed
- —
- Modalities
- text → text
- License
- MIT
Shanghai AI Laboratory's open-weight agentic MoE (744B total, MIT license, 256K context), built on a GLM-5.2 base and trained by the cross-institutional ATRIA initiative (with Fudan NLP Lab) for long-horizon research and engineering workflows: problem analysis, tool use, code, experiments, and failure recovery. Benchmark scores are from the model's own technical report and have not yet been independently verified by a third-party evaluator.
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.