o3 Mini vs Skyfall 36B V2
Skyfall 36B V2 costs 5.5x less per output token than o3 Mini ($0.80 vs $4.40 per million). Updated October 21, 2026; no benchmark has tested both models yet.
- Coding
- No data
- Overall
- 38.0
- Output $/M
- $4.40
- Coding
- No data
- Overall
- No data
- Output $/M
- $0.80
o3 Mini or Skyfall 36B V2: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick o3 Mini for a larger context window (200K against 33K tokens).
No benchmark has tested both yet. o3 Mini has results on Aider polyglot, SWE-bench Verified and SWE-bench Lite; Skyfall 36B V2 has none yet. Compare them on price and context instead.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
No call
Only o3 Mini has a AA Intelligence Index result, 12.5.
Directional only
Reasoning
GPQA Diamond
No call
Only o3 Mini has a GPQA Diamond result, 75.5%.
Directional only
Math
AIME
No call
Only o3 Mini has a AIME result, 86.5%.
Directional only
Knowledge
MMLU Pro
No call
Only o3 Mini has a MMLU Pro result, 78.7%.
Directional only
Long documents
Listed context window
o3 Mini
o3 Mini takes 200K tokens against 33K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Skyfall 36B V2
o3 Mini $0.0033, Skyfall 36B V2 $0.0009 per turn.
List prices
At a glance
- Coding index
- o3 Mini not benchmarked, Skyfall 36B V2 not benchmarked.
- Widest category
- None yet: no category has been tested on both models.
- Widest single gap
- None yet: no benchmark has tested both models.
- Output price
- o3 Mini $4.40 against Skyfall 36B V2 $0.80 per million output tokens, 5.5x apart.
- Context window
- o3 Mini takes 200K tokens, Skyfall 36B V2 33K.
Benchmark by benchmark
Every board either model has a result on, a category to a fold; bars are each result on a 0 to 100 scale. Only one model is measured on each, so no row names a leader.
CodingDirectional onlySix of these boards feed the Vibe Coding Index13 boards measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| Aider polyglot | 60.4% | Not measured | Only o3 Mini measured |
| SWE-bench Verified | 42.4% | Not measured | Only o3 Mini measured |
| SWE-bench Lite | 32.3% | Not measured | Only o3 Mini measured |
| LiveCodeBench (Vals AI) | 71.5% | Not measured | Only o3 Mini measured |
| LiveCodeBench (Artificial Analysis run) | 73.4% | Not measured | Only o3 Mini measured |
| SciCode | 42.8% | Not measured | Only o3 Mini measured |
| Terminal-Bench Hard (Artificial Analysis run) | 6.8% | Not measured | Only o3 Mini measured |
| GSO-Bench | 1.3% | Not measured | Only o3 Mini measured |
| WeirdML | 43.7% | Not measured | Only o3 Mini measured |
| Cybench | 22.5% | Not measured | Only o3 Mini measured |
| LMArena Text: coding | 1378 | Not measured | Only o3 Mini measured |
| LMArena Text: software and IT | 1364 | Not measured | Only o3 Mini measured |
| SciCode | 7.7% | Not measured | Only o3 Mini measured |
| Neither model has a result on the other 26 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| AA Intelligence Index | 12.5 | Not measured | Only o3 Mini measured |
| Epoch Capabilities Index | 140.2 | Not measured | Only o3 Mini measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 31.3% | Not measured | Only o3 Mini measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index12 boards measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| GPQA Diamond | 75.5% | Not measured | Only o3 Mini measured |
| GPQA Diamond (Artificial Analysis run) | 77.3% | Not measured | Only o3 Mini measured |
| Humanity's Last Exam (Artificial Analysis run) | 12.0% | Not measured | Only o3 Mini measured |
| GPQA Diamond (Epoch AI run) | 77.0% | Not measured | Only o3 Mini measured |
| ARC-AGI | 34.5% | Not measured | Only o3 Mini measured |
| ARC-AGI-2 | 3.0% | Not measured | Only o3 Mini measured |
| SimpleBench | 22.8% | Not measured | Only o3 Mini measured |
| Chess puzzles | 15.0% | Not measured | Only o3 Mini measured |
| Mystery game puzzles | 7.0% | Not measured | Only o3 Mini measured |
| LMArena Text: hard prompts | 1365 | Not measured | Only o3 Mini measured |
| LMArena Text: expert | 1364 | Not measured | Only o3 Mini measured |
| LMArena Text: hard prompts in English | 1375 | Not measured | Only o3 Mini measured |
| Neither model has a result on the other 10 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index9 boards measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| AIME | 86.5% | Not measured | Only o3 Mini measured |
| MATH 500 (Vals AI) | 91.8% | Not measured | Only o3 Mini measured |
| MGSM (Vals AI) | 91.3% | Not measured | Only o3 Mini measured |
| OTIS Mock AIME | 76.9% | Not measured | Only o3 Mini measured |
| FrontierMath, tiers 1 to 3 | 18.6% | Not measured | Only o3 Mini measured |
| FrontierMath, tier 4 | 0.0% | Not measured | Only o3 Mini measured |
| MATH level 5 | 96.5% | Not measured | Only o3 Mini measured |
| LMArena Text: math | 1396 | Not measured | Only o3 Mini measured |
| LMArena Text: mathematical work | 1399 | Not measured | Only o3 Mini measured |
| Neither model has a result on the other 4 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index11 boards measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| MMLU Pro | 78.7% | Not measured | Only o3 Mini measured |
| LegalBench (Vals AI) | 71.5% | Not measured | Only o3 Mini measured |
| CorpFin (Vals AI) | 45.3% | Not measured | Only o3 Mini measured |
| TaxEval (Vals AI) | 69.4% | Not measured | Only o3 Mini measured |
| MedQA (Vals AI) | 94.8% | Not measured | Only o3 Mini measured |
| MMLU Pro (Artificial Analysis run) | 80.2% | Not measured | Only o3 Mini measured |
| SimpleQA Verified | 15.3% | Not measured | Only o3 Mini measured |
| LMArena Text: science | 1339 | Not measured | Only o3 Mini measured |
| LMArena Text: medicine and healthcare | 1330 | Not measured | Only o3 Mini measured |
| LMArena Text: legal and government | 1328 | Not measured | Only o3 Mini measured |
| LMArena Text: business and finance | 1313 | Not measured | Only o3 Mini measured |
| Neither model has a result on the other 7 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index23 boards measured
| Benchmark | o3 Mini | Skyfall 36B V2 | Leader |
|---|---|---|---|
| IFBench | 67.1% | Not measured | Only o3 Mini measured |
| Long context reasoning (AA-LCR) | 43.0% | Not measured | Only o3 Mini measured |
| DTBench | 68.8% | Not measured | Only o3 Mini measured |
| LMCA | 19.0% | Not measured | Only o3 Mini measured |
| Fiction.LiveBench | 50.0% | Not measured | Only o3 Mini measured |
| Creative writing (Lech Mazur) | 61.7% | Not measured | Only o3 Mini measured |
| LMArena Text | 1337 | Not measured | Only o3 Mini measured |
| LMArena Text: instruction following | 1337 | Not measured | Only o3 Mini measured |
| LMArena Text: creative writing | 1286 | Not measured | Only o3 Mini measured |
| LMArena Text: English | 1351 | Not measured | Only o3 Mini measured |
| LMArena Text: non-English | 1319 | Not measured | Only o3 Mini measured |
| LMArena Text: Chinese | 1379 | Not measured | Only o3 Mini measured |
| LMArena Text: German | 1303 | Not measured | Only o3 Mini measured |
| LMArena Text: Japanese | 1286 | Not measured | Only o3 Mini measured |
| LMArena Text: Korean | 1313 | Not measured | Only o3 Mini measured |
| LMArena Text: French | 1334 | Not measured | Only o3 Mini measured |
| LMArena Text: Spanish | 1321 | Not measured | Only o3 Mini measured |
| LMArena Text: Russian | 1304 | Not measured | Only o3 Mini measured |
| LMArena Text: Polish | 1294 | Not measured | Only o3 Mini measured |
| LMArena Text: multi-turn | 1320 | Not measured | Only o3 Mini measured |
| LMArena Text: longer queries | 1343 | Not measured | Only o3 Mini measured |
| LMArena Text: writing and language | 1313 | Not measured | Only o3 Mini measured |
| LMArena Text: entertainment and media | 1283 | Not measured | Only o3 Mini measured |
| Neither model has a result on the other 23 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Skyfall 36B V2 costs less
Repository review
50K input + 3K output tokens
Skyfall 36B V2 costs less
Agent loop
200K input + 10K output tokens
Skyfall 36B V2 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | OpenAI o3 Mini | TheDrummer Skyfall 36B V2 |
|---|---|---|
| Maker | OpenAI | TheDrummer |
| Weights | No public weights listed | Open weights |
| Released | January 31, 2025 | March 10, 2025 |
| Context window | 200K | 33K |
| Input $/Mtok | $1.10 | $0.55 |
| Output $/Mtok | $4.40 | $0.80 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is o3 Mini better than Skyfall 36B V2 for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick o3 Mini for a larger context window (200K against 33K tokens).
Which is cheaper, o3 Mini or Skyfall 36B V2?
o3 Mini costs $1.10 per million input tokens and $4.40 per million output tokens; Skyfall 36B V2 costs $0.55 and $0.80. Cost per solved task is known for neither of them.
Which has the bigger context window, o3 Mini or Skyfall 36B V2?
o3 Mini, with 200K tokens against 33K.
Has any benchmark tested both o3 Mini and Skyfall 36B V2?
No benchmark has tested both yet. o3 Mini has results on Aider polyglot, SWE-bench Verified and SWE-bench Lite; Skyfall 36B V2 has none yet. Compare them on price and context instead. Data as of October 21, 2026.