Long documents
Prompts that approach the documented context limit
Nemotron 3 Ultra
Nemotron 3 Ultra has the larger documented context window.
Confidence: documented
Nemotron 3 Ultra vs Ternary Bonsai 2 27B
Updated October 29, 2026. Rank says Ternary Bonsai 2 27B is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Share or export
Decision reading
Ternary Bonsai 2 27B has the higher public point estimate, 53.95 versus 47.37. Their conditional score ranges overlap. These ranges do not establish rank confidence. 7 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 7
- Nemotron 3 Ultra only
- 18
- Ternary Bonsai 2 27B only
- 14
- Like-for-like categories
- 1 / 8
Estimated: Nemotron 3 Ultra and Ternary Bonsai 2 27B. Conditional ranges do not establish rank confidence.How the comparison works
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Show secondary and unsupported calls
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
Ternary Bonsai 2 27B is not ranked on the public lane for coding, so no winner is named for coding.
Confidence: limitedAgentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Ternary Bonsai 2 27B is not ranked on the public lane for agentic, so no winner is named for agentic.
Confidence: limitedChat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-ratesCache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-ratesRepository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Which one for a specific job
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
22.5Nemotron 3 Ultra—Ternary Bonsai 2 27B
Not comparable · BenchAlign v5.8
The coding row is not comparable on the public lane: at least one model is not measured or not ranked there.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Same basis rules as the category table below
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
18 Nemotron 3 Ultra only
7 Shared
14 Ternary Bonsai 2 27B only
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
SWE-bench VerifiedCoding
Normalized gap 11.1Nemotron 3 Ultra:71.9%Ternary Bonsai 2 27B:60.8%IFBenchInstruction following
Normalized gap 7.7Nemotron 3 Ultra:81.7%Ternary Bonsai 2 27B:74%GPQAKnowledge
Normalized gap 1.2Nemotron 3 Ultra:87%Ternary Bonsai 2 27B:85.8%
Bars run 0–100 on each benchmark’s normalized display scale
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.8 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Instruction following
Like-for-like- Nemotron 3 Ultra
- 92.1
- #21/157
- Ternary Bonsai 2 27B
- 74.0
- #73/157
- Basis
- Provisional lane · 1 vs 1 weighted rows
- Reading
- Nemotron 3 Ultra leads
Agentic
Not comparable- Nemotron 3 Ultra
- 25.1
- Supported · #116/135
- Ternary Bonsai 2 27B
- Not ranked
- Basis
- BenchAlign v5.8 lane · 7 vs 3 public rows
- Reading
- Not comparable
Coding
Not comparable- Nemotron 3 Ultra
- 22.5
- Supported · #127/156
- Ternary Bonsai 2 27B
- Not ranked
- Basis
- BenchAlign v5.8 lane · 7 vs 4 public rows
- Reading
- Not comparable
Reasoning
Not comparable- Nemotron 3 Ultra
- 69.5
- Unranked · 1 rankable row
- Ternary Bonsai 2 27B
- 77.9
- Unranked · 1 rankable row
- Basis
- Provisional lane · 1 vs 0 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Nemotron 3 Ultra
- Not ranked
- Ternary Bonsai 2 27B
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Knowledge
Not comparable- Nemotron 3 Ultra
- 50.1
- Estimated · #80/179
- Ternary Bonsai 2 27B
- Not ranked
- Basis
- BenchAlign v5.8 lane · 7 vs 3 public rows
- Reading
- Not comparable
Multilingual
Not comparable- Nemotron 3 Ultra
- 89.2
- #7/17
- Ternary Bonsai 2 27B
- Not ranked
- Basis
- Provisional lane · 1 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- Nemotron 3 Ultra
- Not ranked
- Ternary Bonsai 2 27B
- 86.2
- Unranked · 4 rankable rows
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
| Category | Nemotron 3 Ultra | Ternary Bonsai 2 27B | Basis | Reading |
|---|---|---|---|---|
| Instruction following | 92.1#21/157 | 74.0#73/157 | Like-for-likeProvisional lane · 1 vs 1 weighted rows | Nemotron 3 Ultra leads |
| Agentic | 25.1Supported · #116/135 | Not ranked | Not comparableBenchAlign v5.8 lane · 7 vs 3 public rows | Not comparable |
| Coding | 22.5Supported · #127/156 | Not ranked | Not comparableBenchAlign v5.8 lane · 7 vs 4 public rows | Not comparable |
| Reasoning | 69.5Unranked · 1 rankable row | 77.9Unranked · 1 rankable row | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Multimodal | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Knowledge | 50.1Estimated · #80/179 | Not ranked | Not comparableBenchAlign v5.8 lane · 7 vs 3 public rows | Not comparable |
| Multilingual | 89.2#7/17 | Not ranked | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Math | Not ranked | 86.2Unranked · 4 rankable rows | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Supported evidence per lane · bars run 0–100Methodology
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- Nemotron 3 Ultra
- Self-hosted; infrastructure cost varies
- Fits in one request
- Ternary Bonsai 2 27B
- Self-hosted; infrastructure cost varies
- Fits in one request
Nemotron 3 Ultra has no comparable published API token rate. Ternary Bonsai 2 27B has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Nemotron 3 Ultra
- Self-hosted; infrastructure cost varies
- Fits in one request
- Ternary Bonsai 2 27B
- Self-hosted; infrastructure cost varies
- Fits in one request
Nemotron 3 Ultra has no comparable published API token rate. Ternary Bonsai 2 27B has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Nemotron 3 Ultra
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
- Ternary Bonsai 2 27B
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
Nemotron 3 Ultra has no comparable published API token rate. Ternary Bonsai 2 27B has no comparable published API token rate.
Cached input falls back to the list input rate only where a cached rate is unpublished
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationNemotron 3 UltraTernary Bonsai 2 27B
Context window
Maximum documented context; output-token limits may be lower.
Nemotron 3 Ultra
1M
Ternary Bonsai 2 27B
API model ID
Nemotron 3 Ultra
Not sourced
Ternary Bonsai 2 27B
prism-ml/Ternary-Bonsai-2-27B-gguf
PrismML Bonsai 2 collectionCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Nemotron 3 Ultra
No comparable hosted API rate
Ternary Bonsai 2 27B
No comparable hosted API rate
PrismML Bonsai 2 collectionDocumented inputs
Nemotron 3 Ultra
Not sourced
Ternary Bonsai 2 27B
Not sourced
Documented outputs
Nemotron 3 Ultra
Not sourced
Ternary Bonsai 2 27B
Not sourced
Provider availability
Nemotron 3 Ultra
Not sourced
Ternary Bonsai 2 27B
Not sourced
Reasoning profile
Nemotron 3 Ultra
Reasoning
Ternary Bonsai 2 27B
Reasoning
Weight access
Nemotron 3 Ultra
Open Weight
Ternary Bonsai 2 27B
Open Weight
License
Nemotron 3 Ultra
Open Weight
Ternary Bonsai 2 27B
Open Weight
Release date
Nemotron 3 Ultra
2026-06-04
Ternary Bonsai 2 27B
2026-09-17
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- Ternary Bonsai 2 27B has the higher public point estimate, 53.95 versus 47.37. Their conditional score ranges overlap. These ranges do not establish rank confidence.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Nemotron 3 Ultra has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Questions
Which is better, Nemotron 3 Ultra or Ternary Bonsai 2 27B?
Ternary Bonsai 2 27B has the higher public point estimate, 53.95 versus 47.37. Their conditional score ranges overlap. These ranges do not establish rank confidence. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, Nemotron 3 Ultra or Ternary Bonsai 2 27B?
Ternary Bonsai 2 27B is not ranked on the public lane for coding, so no winner is named for coding.
Which is better for agentic tasks, Nemotron 3 Ultra or Ternary Bonsai 2 27B?
Ternary Bonsai 2 27B is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.
Which costs less, Nemotron 3 Ultra or Ternary Bonsai 2 27B?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, Nemotron 3 Ultra or Ternary Bonsai 2 27B?
Nemotron 3 Ultra has the larger documented context window: 1M, compared with 262K.
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
Nemotron 3 Ultra
API / mo$0
Self-host / mo$7,008
Break-even—
Ternary Bonsai 2 27B
API / mo$0
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence39 rows
Agentic
Terminal-Bench 2.1
Nemotron 3 Ultra56.4%SourceTernary Bonsai 2 27B52.8%SourceNemotron 3 Ultra leads this result
PinchBench
Nemotron 3 Ultra90.0%SourceTernary Bonsai 2 27B—Not directly comparable
BrowseComp
Nemotron 3 Ultra44.4%SourceTernary Bonsai 2 27B—Not directly comparable
τ³-bench results
Nemotron 3 Ultra70.9%SourceTernary Bonsai 2 27B—Not directly comparable
HLE w/ tools
Nemotron 3 Ultra37.4%SourceTernary Bonsai 2 27B—Not directly comparable
Terminal-Bench 2.1 (Vals)
Nemotron 3 Ultra50.9%SourceTernary Bonsai 2 27B—Not directly comparable
MazeBench
Nemotron 3 Ultra0%SourceTernary Bonsai 2 27B—Not directly comparable
τ²-bench results
Nemotron 3 Ultra—Ternary Bonsai 2 27B80.2%SourceNot directly comparable
BFCL v3
Nemotron 3 Ultra—Ternary Bonsai 2 27B74.9%SourceNot directly comparable
Coding
SWE-bench Verified
Nemotron 3 Ultra71.9%SourceTernary Bonsai 2 27B60.8%SourceNemotron 3 Ultra leads this result
SWE Multilingual
Nemotron 3 Ultra67.7%SourceTernary Bonsai 2 27B—Not directly comparable
LiveCodeBench v6
Nemotron 3 Ultra89.0%SourceTernary Bonsai 2 27B90.1%SourceTernary Bonsai 2 27B leads this result
SciCode
Nemotron 3 Ultra44.6%SourceTernary Bonsai 2 27B—Not directly comparable
Terminal-Bench 2.1
Nemotron 3 Ultra56.4%SourceTernary Bonsai 2 27B52.8%SourceNemotron 3 Ultra leads this result
LiveCodeBench (Vals)
Nemotron 3 Ultra86.0%SourceTernary Bonsai 2 27B—Not directly comparable
SWE-bench (Vals)
Nemotron 3 Ultra69.0%SourceTernary Bonsai 2 27B—Not directly comparable
BigCodeBench
Nemotron 3 Ultra—Ternary Bonsai 2 27B58.1%SourceNot directly comparable
Reasoning
Multimodal
CharXiv (overall)
Nemotron 3 Ultra—Ternary Bonsai 2 27B80.0%SourceNot directly comparable
A-OKVQA
Nemotron 3 Ultra—Ternary Bonsai 2 27B86.8%SourceNot directly comparable
OmniDocBench 1.6
Nemotron 3 Ultra—Ternary Bonsai 2 27B89.1%SourceNot directly comparable
RealWorldQA
Nemotron 3 Ultra—Ternary Bonsai 2 27B80.1%SourceNot directly comparable
OCRBench V2
Nemotron 3 Ultra—Ternary Bonsai 2 27B56.9%SourceNot directly comparable
Knowledge
GPQA
Nemotron 3 Ultra87%SourceTernary Bonsai 2 27B85.8%SourceNemotron 3 Ultra leads this result
GPQA-D
Nemotron 3 Ultra87.0%SourceTernary Bonsai 2 27B85.8%SourceNemotron 3 Ultra leads this result
HLE
Nemotron 3 Ultra26.7%SourceTernary Bonsai 2 27B—Not directly comparable
HLE w/o tools
Nemotron 3 Ultra26.7%SourceTernary Bonsai 2 27B—Not directly comparable
MMLU-Pro
Nemotron 3 Ultra86.8%SourceTernary Bonsai 2 27B—Not directly comparable
GPQA Diamond (Vals)
Nemotron 3 Ultra86.1%SourceTernary Bonsai 2 27B—Not directly comparable
MMLU-Pro (Vals)
Nemotron 3 Ultra85.8%SourceTernary Bonsai 2 27B—Not directly comparable
MMLU-Redux
Nemotron 3 Ultra—Ternary Bonsai 2 27B89.1%SourceNot directly comparable
Multilingual
MMLU-ProX
Nemotron 3 Ultra83%SourceTernary Bonsai 2 27B—Not directly comparable
Instruction following
Math
GSM8K
Nemotron 3 Ultra—Ternary Bonsai 2 27B96.7%SourceNot directly comparable
MATH-500
Nemotron 3 Ultra—Ternary Bonsai 2 27B98.8%SourceNot directly comparable
AIME 2025
Nemotron 3 Ultra—Ternary Bonsai 2 27B95%SourceNot directly comparable
AIME26
Nemotron 3 Ultra—Ternary Bonsai 2 27B95.8%SourceNot directly comparable
39 public results · 7 shared
Watch Nemotron 3 Ultra vs Ternary Bonsai 2 27B
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 6,100+ subscribers.
Last updated October 29, 2026