Model comparison
Deepseek Coder v2 vs Inkling
Inkling is the stronger model overall, scoring 44.1 to 35.3 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 2 categories and Inkling in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Inkling leads 65.5 to 38.4.
Side by side
| Deepseek Coder v2 | Inkling | |
|---|---|---|
| Provider | DeepSeek | Thinking Machines Lab |
| Noometry Index | 35.3 | 44.1 |
| Released | 2024-06-17 | 2026-07-15 |
| Weights | Open | Open |
| Context window | — | 66K |
| Max output | — | 66K |
| Input $ / M tokens | — | $1.87 |
| Output $ / M tokens | — | $4.68 |
| Results tracked | 24 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Deepseek Coder v2 leads
Deepseek Coder v2: 37.5 (#178), Inkling: 34.1 (#229)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Coding | 1251 | 1463 |
| FrontierCode | — | 14% |
| LMArena WebDev | — | 1412 |
| FrontierSWE | — | 4.1% |
| SciCode | — | 47% |
| WeirdML | — | 32.3% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| ALE-Bench | — | 946 |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Agentic & Tool Use Not comparable
Deepseek Coder v2: —, Inkling: 29.5 (#76)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| τ²-bench Banking | — | 25% |
Reasoning Inkling leads
Deepseek Coder v2: 22.6 (#175), Inkling: 40.6 (#48)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1452 |
| ARC-AGI-2 | — | 36.5% |
| SimpleBench | — | 50% |
| ARC-AGI-1 | — | 79.5% |
| CritPt | — | 5.4% |
| Chess Puzzles | — | 21% |
| DTBench | — | 87.5% |
| LMCA | — | 37.6% |
| Epoch Capabilities Index | — | 148.6 |
| WinoGrande | 83.7% | — |
Math Deepseek Coder v2 leads
Deepseek Coder v2: 32.3 (#188), Inkling: 31.2 (#198)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Math | 1241 | 1482 |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| FrontierMath Tier 4 | — | 4.9% |
| OTIS Mock AIME 2024-2025 | — | 88.9% |
| ProofBench | — | 0% |
| GSM8K | 94.5% | — |
Knowledge Inkling leads
Deepseek Coder v2: 31.7 (#209), Inkling: 55.2 (#44)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Expert | 1181 | 1468 |
| GPQA Diamond | — | 88.3% |
| SimpleQA Verified | — | 40.3% |
| ARC (AI2) Challenge | 64.3% | — |
Multilingual Inkling leads
Deepseek Coder v2: 36.4 (#235), Inkling: 54.2 (#49)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Non-English | 1182 | 1434 |
| LMArena Chinese | 1201 | 1495 |
| LMArena French | 1184 | 1453 |
| LMArena German | 1164 | 1456 |
| LMArena Japanese | 1126 | 1428 |
| LMArena Korean | 1104 | 1401 |
| LMArena Russian | 1188 | 1430 |
| LMArena Spanish | 1153 | 1442 |
Instruction Following Inkling leads
Deepseek Coder v2: 61.9 (#237), Inkling: 75.3 (#65)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Instruction Following | 1180 | 1427 |
Long Context Inkling leads
Deepseek Coder v2: 37.0 (#219), Inkling: 43.9 (#79)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Longer Query | 1219 | 1436 |
Writing & Preference Inkling leads
Deepseek Coder v2: 38.4 (#249), Inkling: 65.5 (#48)
| Benchmark | Deepseek Coder v2 | Inkling |
|---|---|---|
| LMArena Text | 1191 | 1443 |
| LMArena Creative Writing | 1120 | 1390 |
| LMArena Multi-Turn | 1176 | 1436 |
| EQ-Bench Creative Writing | — | 1611 |
| EQ-Bench 4 | — | 1226 |
Frequently asked questions
Is Deepseek Coder v2 better than Inkling?
Inkling is the stronger model overall, scoring 44.1 to 35.3 on the Noometry Index.
Is Deepseek Coder v2 or Inkling better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 37.5 versus 34.1 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and Inkling share?
17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Inkling has 41.