DeepReinforceLanguage model
Ornith 1.0 397B
Ornith 1.0 397B has no overall index value yet because there are too few independent benchmark results.
PopularityRank 320 of 785
39out of 100-6 vs. last month
based on search, usage and attention
Summary
Ornith 1.0 397B is the large MoE tier in the first Ornith family, with 397 billion total parameters. The checkpoint accepts image and text input with a 262K context. The defining training idea remains joint reinforcement-learning improvement of solution paths and agent scaffolds.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 19 published benchmark results
Page 1 of 2
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
Terminal-Bench 2.1 (Terminus-2) (Show measurement, test conditions, and source)
- Source value
- 77.5
- Score
- 77.5
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- Terminal-Bench 2.1; Terminus-2; five runs
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B table row; Terminal-Bench 2.1 using Terminus-2 with parser=json, temperature=1, top_p=1, 128k context, 4h timeout, 32 CPU/48GB, five runs.
- Context
- Publisher-reported values; exact task revisions are not fully pinned.
Terminal-Bench 2.1 (Claude Code 2.1.126) (Show measurement, test conditions, and source)
- Source value
- 78.2
- Score
- 78.2
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- Terminal-Bench 2.1; Claude Code 2.1.126; five-run average
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B table row; Terminal-Bench 2.1 using Terminus-2 with parser=json, temperature=1, top_p=1, 128k context, 4h timeout, 32 CPU/48GB, five runs.
- Context
- Publisher-reported values; exact task revisions are not fully pinned.
SWE-Bench Verified (Show measurement, test conditions, and source)
- Source value
- 82.4
- Score
- 82.4
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- SWE-Bench Verified; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
SWE-Bench Pro (Show measurement, test conditions, and source)
- Source value
- 62.2
- Score
- 62.2
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- SWE-Bench Pro; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
SWE-Bench Multilingual (Show measurement, test conditions, and source)
- Source value
- 78.9
- Score
- 78.9
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- SWE-Bench Multilingual; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
NL2Repo (Show measurement, test conditions, and source)
- Source value
- 48.2
- Score
- 48.2
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- NL2Repo; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
ClawEval (Show measurement, test conditions, and source)
- Source value
- 77.1
- Score
- 77.1
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- ClawEval; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
SWE Atlas QnA (Show measurement, test conditions, and source)
- Source value
- 41.2
- Score
- 41.2
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- SWE Atlas QnA; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
SWE Atlas RF (Show measurement, test conditions, and source)
- Source value
- 42.6
- Score
- 42.6
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- SWE Atlas RF; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
SWE Atlas TW (Show measurement, test conditions, and source)
- Source value
- 39.1
- Score
- 39.1
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- SWE Atlas TW; exact harness/sampling not stated
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- DeepReinforce Ornith 1.0 397B model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Ornith-1.0-397B score row; the card reports these benchmark names but does not fully specify the harness, decoding parameters or task snapshot for each.
- Context
- The publisher does not disclose a complete per-row evaluation protocol; raw scores are not normalized.
DeepSWE (Ornith-1.0-397B) (Show measurement, test conditions, and source)
- Source value
- 8
- Score
- 8
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- Ornith-1.0 benchmark table, pinned repository README commit 9c800446621ebad1172f517a1d0f9b683fb8615e
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Ornith Team, Ornith-1 repository benchmark table
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Vendor-reported result. The exact source-table column is Ornith-1.0-397B. The pinned README checkpoint list links this exact 1.0 checkpoint to ornith-ai/Ornith-1.0-397B. The profile's Hugging Face API route resolved to canonical model ID ornith-ai/Ornith-1.0-397B at current revision 5e3e761811e804c295c1d3c0ce68b21da6154209; this confirms route identity, not the checkpoint revision used in the reported evaluation. Benchmark-specific setup: For DeepSWE, the README specifies the Claude Code harness, temperature 1.0, top_p 0.95 and a 256K context window. The table does not specify whether this 1.0 cell used the same run aggregation; no model serving route or per-cell N is stated.
- Context
- Vendor-reported score. The table labels no numeric unit and gives no cell-specific N or confidence interval. The tested serving route and exact evaluation checkpoint revision are not pinned; no percentage unit or sample size is inferred.
Frontier-Bench v0.1 (Ornith-1.0-397B) (Show measurement, test conditions, and source)
- Source value
- 2.7
- Score
- 2.7
- Metric
- score as published
- Unit
- No unit provided
- Benchmark version
- Ornith-1.0 benchmark table, pinned repository README commit 9c800446621ebad1172f517a1d0f9b683fb8615e
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Ornith Team, Ornith-1 repository benchmark table
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Vendor-reported result. The exact source-table column is Ornith-1.0-397B. The pinned README checkpoint list links this exact 1.0 checkpoint to ornith-ai/Ornith-1.0-397B. The profile's Hugging Face API route resolved to canonical model ID ornith-ai/Ornith-1.0-397B at current revision 5e3e761811e804c295c1d3c0ce68b21da6154209; this confirms route identity, not the checkpoint revision used in the reported evaluation. Benchmark-specific setup: The README names this benchmark but does not specify the runner, decoding setup or denominator. The table does not specify whether this 1.0 cell used the same run aggregation; no model serving route or per-cell N is stated.
- Context
- Vendor-reported score. The table labels no numeric unit and gives no cell-specific N or confidence interval. The tested serving route and exact evaluation checkpoint revision are not pinned; no percentage unit or sample size is inferred.
Page 1 of 2
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 101 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 7, 2026
Ornith 1.0 397BAliceAI Foundation 80B A3B Base 39Command R7B 39Devstral Small 2 39Ling 3.0 Flash 39
Method changes in the timeline, a jump at these points is no market event:
- 1November 1, 2026: Value only with search or usage data, 30-day attention
What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.
Similar models
More models of the same kind as Ornith 1.0 397B, chosen by model line, developer, size, price and release date.
- Ornith 1.0 35BDeepReinforce · same developer
- Ornith 1.0 9BDeepReinforce · same developer
- Ornith 1.5 35B A3BDeepReinforce · same developer
- Ornith 1.0 31BDeepReinforce · same developer
- Nex-N2-ProNex AGI · similar size
- K2 Horizon 375B-A23BInstitute of Foundation Models · similar size
- Qwen 3.5 397B A17BAlibaba · similar size
- Trinity Large ThinkingArcee AI · similar size
All specifications
Architecture
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Frequently asked questions about Ornith 1.0 397B
How large is the context window of Ornith 1.0 397B?
262,144 tokens. That is how much text Ornith 1.0 397B reads at once, about 390 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Ornith 1.0 397BRank 67 of 103262K
Kimi K2.6262K
Qwen 3.8 27B262K
Kimi K2.7 Code262K
Kimi K2.5262K
Nemotron 3 Ultra262K
Kimi K2 Thinking262K
Can I run Ornith 1.0 397B locally?
Yes, the weights are open. With 397B parameters it needs a lot of memory, more a server than a laptop.
Stated by DeepReinforce
- Open weights
- yes
- License
- MIT
- Parameters
- 397B
- active per token
- n/a
Sources and data date
Show all 12 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Input | Documented modalities. Retrieved October 27, 2026. |
| Output | Documented modalities. Retrieved October 27, 2026. |
| Research status | Research date October 15, 2026. 3 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not published | benchmark-units-and-snapshots Note Most table rows do not publish units, exact task revisions or complete harness settings; the source values are preserved as unknown-unit raw scores. |
| Additional source | Hugging Face · Ornith 1.0 397B model card · Editorial description reviewed October 2, 2026 |
| Additional source | DeepReinforce Ornith 1.0 397B model card (retrieved October 15, 2026) |
| Additional source | Ornith Team, Ornith-1 repository benchmark table (retrieved October 18, 2026) |
| Additional source | Model metadata (retrieved October 18, 2026) |
| Additional source | Ornith 1.0 family model card · Editorial description reviewed October 2, 2026 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |