vs
2:6
overall capability wins (0 ties across 8 scored). Mercury 2 leads overall.
- Reasoning1:5
- Coding0:1
- Pricing3:0
- Specs1:0
Verdict
Mercury 2 leads coding (widest gap: +0.7 on SciCode); Ling-2.6 1T offers ~2.0x larger context (262K vs 128K); Ling-2.6 1T ships open weights (MIT).
- Mercury 2 leads coding (widest gap: +0.7 on SciCode)
- Ling-2.6 1T offers ~2.0x larger context (262K vs 128K)
- Ling-2.6 1T ships open weights (MIT)
- Ling-2.6 1T is ~1.8x cheaper on a blended token basis than Mercury 2
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Ling-2.6 1T
- $0.23
- Mercury 2
- $0.44
Ling-2.6 1T is about 1.9x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
InclusionAI
Inception Labs
Family
Other
Other
License
MIT
Proprietary
Open weights
Yes
No
Release
Mar 10, 2026
Sep 1, 2026
Knowledge cutoff
-
-
API / provider
InclusionAI
Inception Labs
Modalities
text → text
text → text
Specs
Context window
262K
128K
Max output
33K
-
Parameters
—
—
Pricing
Input $/1M
$0.07
$0.25
Output $/1M
$0.63
$0.75
Blended $/1M (3∶1)
$0.21
$0.38
Speed
tok/s
55
-
TTFT (s)
0.55
-
Reasoning
MMLU-Pro
—
—
GPQA Diamond
75.2
77.0
Humanity's Last Exam
8.7
17.1
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
21.9
21.2
AA-LCR v1.1
41.7
43.7
CritPt
0.3
0.8
MMMU-Pro
—
—
IFBench
56.9
69.8
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
—
27.3
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
37.0
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 | Ling-2.6 1T | Mercury 2 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | InclusionAI | Inception Labs | — |
| Family | Other | Other | — |
| License | MIT | Proprietary | — |
| Open weights | Yes | No | — |
| Release | Mar 10, 2026 | Sep 1, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | InclusionAI | Inception Labs | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 262K | 128K | A +134K |
| Max output | 33K | - | — |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $0.07 | $0.25 | A +$0.17 |
| Output $/1M | $0.63 | $0.75 | A +$0.13 |
| Blended $/1M (3∶1) | $0.21 | $0.38 | A +$0.16 |
| Speed | |||
| tok/s | 55 | - | — |
| TTFT (s) | 0.55 | - | — |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 75.2 | 77.0 | B +1.8 pts |
| Humanity's Last Exam | 8.7 | 17.1 | B +8.4 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 21.9 | 21.2 | A +0.7 pts |
| AA-LCR v1.1 | 41.7 | 43.7 | B +2.0 pts |
| CritPt | 0.3 | 0.8 | B +0.5 pts |
| MMMU-Pro | — | — | — |
| IFBench | 56.9 | 69.8 | B +12.9 pts |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | — | 27.3 | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 37.0 | 37.7 | B +0.7 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 | — | — | — |
Ling-2.6 1T
- Context
- 262K / 33K out
- Parameters
- —
- Price
- $0.07 / $0.63
- Speed
- 55 tok/s · 0.55s TTFT
- Modalities
- text → text
- License
- MIT
InclusionAI trillion-scale flagship — stronger reasoning sibling to Ling Flash.
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.