The story so far
Weights, every dated article, and repo commits on one rail.
18 repo commits since the weights landed
Teal is the weights (2026-09-26), grey ticks are repo commits.
jebadiah-27b
frontier-infra Apache 2.0 out 29 days ago
27.8B parameters, images and text in, text out, 256K tokens of context.
Verified 2026-10-25 Rank #0 in Decision models · score 0 method v1 Sources 3 Downloads 30d 638 · #603 of 619 tracked
Do people use it?638downloads, 30 days#603 of 619 we track · below the median+0 likes, downloads +16% this week4 likesHugging Face2026-10-24
Where it's used
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 5 quantized builds and 5 Spaces.
Will it fit?
Weights only, an estimate: leave room for the KV cache. Method
Fits a 64 GB Mac (unified memory) at INT4
About 17 GB at INT4, 33 GB at FP8, 67 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
What people built on it
5 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
5 derivatives on the Hub, most of them quantized
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
Benchmarks
8 rows
Benchmarks 8 rows
measured third-party leaderboard rowauthor-reported number from the model card or eval files (8)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported Banking77 · Classify and make typed decisions | 77 percent | 27B | model author | 2026-10-23 |
| author-reported Decision Index (DI) · Classify and make typed decisions | 54.67 | Jebadiah 27B | model author | 2026-10-23 |
| author-reported JevBench v1.4.2 | 0.87 | Jebadiah 27B | model author | 2026-10-23 |
| author-reported Jevals HelpSteer2 helpfulness (score, 300) | 48.7 percent | 27B | model author | 2026-10-23 |
| author-reported Jevals PubMedQA (noul, 300) | 90 percent | 27B | model author | 2026-10-23 |
| author-reported Kev transfer-v4 test (mixed, 764) | 85.9 percent | 27B | model author | 2026-10-23 |
| author-reported Nimble held-out eval (mixed, 324) | 93.5 percent | 27B | model author | 2026-10-23 |
| author-reported Nimble public, 13 subsets (macro, 3,880) | 78.6 percent | 27B | model author | 2026-10-23 |
Use-case evidence
6 use cases
What it is good at 6 use cases
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
- Classify and make typed decisionsauthor-reportedAuthor reports Decision Index 0.2.1 54.67Author reports Jevals Banking77 (choice, 77 options, 300) 77.0decision-model estimate p=1.00+2 more
- Understand images and videostructuraltag: image-text-to-textpipeline: image-text-to-textconfig: vision_config
- Chat and follow instructionsstructuraltag: conversationalchat template present
- Call tools and APIsstructuralchat template: <tool_call>
- Reason through hard problemsstructuralchat template: <think>
- Read long documentsstructuralconfig: 262,144 context
Facts
20 fields
Listed in Decision models
Facts
- jebadiah-27b has 27.8B parameters. (safetensors metadata, observed 2026-10-24)
- Context length is 262,144 tokens (256K). (config.json max_position_embeddings, observed 2026-10-24)
- License: Apache 2.0 (friction: none). Permissive; patent grant. (license file, observed 2026-10-24)
- Created 2026-09-26, last modified 2026-10-16. (HF API, observed 2026-10-24)
| Pipeline tag | image-text-to-text |
|---|---|
| Library | transformers |
| Architectures | Qwen3_5ForConditionalGeneration |
| Model type | qwen3_5 |
| Dtype | bfloat16 |
| Params by dtype | BF16: 27,781,427,952 |
| Size band | L |
| MoE | no |
| Context length | 262,144 (config.json max_position_embeddings) |
| Languages | en |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
Servability
vLLM · SGLang
Servability
vLLM: supported · Qwen3_5ForConditionalGeneration · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Qwen3_5ForConditionalGeneration · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Hardware fit
3 estimates
Hardware fit ESTIMATE
Weights-only memory for jebadiah-27b: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 67 GB · Single high-end GPU (<80 GB) |
|---|---|
| FP8 (1 byte/param) | 33 GB · Single high-end GPU (<80 GB) |
| INT4 (0.5 bytes/param) | 17 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
Decision facts
catalog
Decision-model catalog curated
| How it was made | LoRA/adapter fine-tune |
|---|---|
| Lineage bucket | F - fine-tune of an open model (public/synthetic data) |
| Lineage evidence | board file inspection (adapter/head/weight files) + HF base_model field |
| Base model | Qwen/Qwen3.8-27B |
| DI 0.2.1 score | 54.67 (independent (DI 0.2.1 board)) |
| ECE (board) | 0.0140 |
| Median latency | 110.4 ms (DI board, 1x RTX PRO 6000, single process) |
| Hosted price (input) | self-hosted/M |
| Source | https://huggingface.co/frontier-infra/jebadiah-27b |
All variants
5 variants
Lineage
Base model: Qwen/Qwen3.8-27B
quantized: 5
Top variants (5 shown, ranked by downloads)
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| mradermacher/jebadiah-27b-GGUF | quantized | GGUF | — | — | apache-2.0 | mradermacher | 1 | 1.9k |
| frontier-infra/jebadiah-27b-GGUF | quantized | GGUF | — | — | apache-2.0 | frontier-infra | 0 | 1.4k |
| frontier-infra/jebadiah-27b-MLX-4bit | quantized | MLX | 26.9B | L | apache-2.0 | frontier-infra | 0 | 119 |
| frontier-infra/jebadiah-27b-MLX-8bit | quantized | MLX | 26.9B | L | apache-2.0 | frontier-infra | 0 | 56 |
| frontier-infra/jebadiah-27b-MLX | quantized | MLX | — | — | apache-2.0 | frontier-infra | 0 | 0 |
Repository changelog
18 commits
Repository changelog 18 commits
| Date | Commit | SHA |
|---|---|---|
| 2026-10-16 | Link the playground from the start-here block | c44f505 |
| 2026-10-16 | Add the start-here block and move build details down | 844f79f |
| 2026-10-10 | Add local client install and project links | 9d236f5 |
| 2026-10-08 | Card: note that images are untested (vision tower inherited, trained and calibrated on text only) | dfa3697 |
| 2026-10-04 | Add the agent support line | 4d3ed81 |
| 2026-10-03 | Update the Decision Index line and add curated-list listings | a2d1031 |
| 2026-09-29 | Card: 9B v2 and 4B v2 comparisons at their refit temperatures | 3dd6f22 |
| 2026-09-29 | Link the Run Jeb locally guide from Which one should I use | 54b89f6 |
| 2026-09-28 | Model card: results links and community quants | 6bc2299 |
| 2026-09-26 | Card: the server has run the 27B, AINode 0.5.32 and 0.5.33 are released, add the JevBench public-items result | c68db2b |
| 2026-09-26 | Card: point "Run it anywhere" to the LM Studio instructions | 1c22cdd |
| 2026-09-26 | Card: link the GGUF and MLX builds | eb78550 |
| 2026-09-26 | Temperatures: apply the hard fit to score questions (0.76), keep train for choice and noul | 1c0d794 |
| 2026-09-26 | Card: drop the private-repo login note | a5d7c80 |
| 2026-09-26 | Model card: two run sections (anywhere, AINode), compact results; full tables to eval/RESULTS.md | 4480932 |
| 2026-09-26 | Jebadiah 27B: model card | 519f0b3 |
| 2026-09-26 | Jebadiah 27B: merged bf16 weights, tokenizer, temperatures, scripts, eval records | 6690054 |
| 2026-09-26 | initial commit | b0e0eb7 |
Similar models
3 models
Similar models
- jpt-9b · 9.41B same size band (L), same task (image-text-to-text)
- Eikos-27B-FP8 · 27.8B same size band (L), permissive license
- openjev · 27.4B same size band (L)
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:44 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:19 UTC · Last verified 2026-10-25
Follow jebadiah-27b
Watch this model
- /feeds/models/frontier-infra--jebadiah-27b.xml(RSS 2.0) · every change to jebadiah-27b
- /feeds/families/frontier-infra--jebadiah.xml(RSS 2.0) · every change across the jebadiah family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.