Tev1-4B-experimental
Quadrant scores
See the full quadrantScored with the public quadrant rubric (maturity × capability; bubble = adoption). Revise as evidence lands.
- Maturity5.8/10
Open weights, full training recipe, hosted serverless endpoint, and public inference examples; still experimental, generative, and without a published latency benchmark or SLA.
- Capability2.7/10
Decision Index 0.4.1 Full 33.73 (#143). Choice-only output from a 2–24 option prompt; Qwen's autoregressive head, so token logprobs are preferences, not calibrated probabilities; no latency measured.
- Adoption19/100
4B: ~2.4k HF downloads and 24 likes (was 741 / 12); 0.8B: ~800 downloads; GitHub recipe 177 stars. Since 30 Sep both sizes ship in Ollama (~2.7k pulls in ~14h). Still an experimental niche release.
Vendor claims
- 88.0% on 1,000 main development decisions and 300/300 policy-transfer decisions[Vendor claim — not independently verified]
- Training the recipe costs about $17[Vendor claim — not independently verified]
- Serverless pricing of $0.04 per million input tokens (together.ai/pricing, 9 Oct 2026; was $0.042) and $0 per million output tokens[Vendor claim — not independently verified]
Board audit of 17 Oct 2026: see the “Board audit” section below; where it conflicts with older board statements on this page, it wins.
Tev1-4B-experimental is an open, Jev-inspired decision model from Together AI. It is a supervised fine-tune of Qwen3.5-4B trained to choose one option from a structured state, question, and list of 2–24 choices.
The model card is explicit about the boundary: this is not a non-autoregressive Jev runtime. It keeps Qwen’s standard language-model head, and the application maps the returned option letter back to a semantic key. For that reason, the catalog entry is limited to choice; it does not claim Jev’s calibrated Score or Boolean interfaces.
Training recipe
The public repository includes the data recipe and a training example. The new v1 run combines 37,840 unique training examples with 4,568 validation examples across language classification, policy decisions, routing, and synthetic research classification. The training example uses LoRA-style supervised fine-tuning on Qwen3.5-4B.
The repository warns that the saved evaluation mixture influenced model development. The reported 88.0% result on 1,000 main decisions and 300/300 policy-transfer decisions are bring-up results, not an independent benchmark. The public model card also says that logprobs are model preferences, not calibrated confidence.
The 0.8B sibling, Tev1-0.8B-experimental, went public on Hugging Face on 25 September 2026. Same recipe and interface on Qwen3.5-0.8B; its card says the weight licence “is being finalized”. It is covered here rather than as a separate entry.
On Ollama (30 Sep 2026)
Both sizes are launch models for Ollama’s /v1/systemone endpoint: ollama pull tev1 (4B, 4.5 GB Q8_0, 4.21B params) and ollama pull tev1:0.8b (812 MB Q8_0, 752M params). Ollama’s library page lists three question types (choice, noul, score) and advises staying within 2–24 options, the range Tev1 was trained on.
The catalog keeps decisionTypes at choice. On Ollama, the noul and score shapes come from Ollama’s candidate scoring, not from Tev1’s training, and Together’s card still describes a one-letter output with uncalibrated logprobs.
UnverifiedClaim: Ollama’s run on Bespoke Labs’ 13 public human-labelled datasets (3,880 decisions) gives 73.3% for Tev1 4B and 63.5% for Tev1 0.8B, against 76.0% for Jev 1.13 (Jev figure from Bespoke’s published run). (blog · library)
Fit / anti-fit
Fit when you need: an open Jev-inspired classifier, a reproducible fine-tuning recipe, or a controlled choice decision behind a typed application schema.
Anti-fit when you need: a non-autoregressive System One runtime, calibrated probabilities, Boolean/Noul or Score semantics, or production decisions without your own held-out evaluation.
Limits
- No public p50/p99 latency benchmark. The
0–0msfrontmatter value is an unavailable-data sentinel. - The 88% and 100% figures are development results and UnverifiedClaim.
- The model is still a generative Qwen checkpoint and may produce prose outside the constrained interface.
- The
$0.04/Minput and$0/Moutput prices are vendor claims for Together’s serverless endpoint.
Board audit (17 Oct 2026)
Checked against the official boards on 17 Oct 2026: Decision Index 0.3.2 (data of 16 Oct, 19:00 UTC), JevBench and ImageJevBench live. The boards are separate sources and are never averaged. Older board statements and rankings further down this page are superseded where they conflict with this section.
| Board | Row | Result |
|---|---|---|
| Decision Index 0.3.2 (16 Oct 19:00 UTC) | Together Tev1 Qwen3.5-4B | Full 34.14 (#112, tie group 28); public 29.51 (#129); same-skills 31.78; new-domains 37.65 |
| Decision Index 0.3.2 (16 Oct 19:00 UTC) | Together Tev1 Qwen3.5-0.8B | Full 12.02 (#149, tie group 45); public 13.1 (#157); same-skills 12.84; new-domains 14.73 |
Score proposal (from scratch, audit follow-up 17 Oct 2026). Capability 2.3 → 2.7 = types × 0.30 + calibration × 0.25 + latency × 0.25 + accuracy × 0.20 = 3 × 0.30 + 1 × 0.25 + 2 × 0.25 + 5 × 0.20 = 2.65 → 2.7. Types 3 (choice only). Calibration 1 and latency 2 are unchanged from the stored working (uncalibrated logprobs; latency not measured). Accuracy 3 → 5: the earlier 3 rested on the author’s development benchmark; on the board, Decision Index Full 34.14 ÷ 60.11 × 8 = 4.5 → 5. The 0.8B sibling (Full 12.02, #149) is covered here but not scored separately. Maturity and adoption unchanged.
Hugging Face on 17 Oct (togethercomputer/Tev1-4B-experimental): 5,096 downloads, 36 likes, licence tag none, last modified 2026-09-23.
Board audit (20 Oct 2026, DI 0.4)
Decision Index 0.4 (20 Oct; Full now from the private parts only) lists Together Tev1 Qwen3.5-4B at Full 33.60 (#127); it was 34.14 on the archived 0.3.2 edition. Accuracy rule, reference Jev 1.13 on DI 0.4 Full (61.26): 8 × 33.60 ÷ 61.26 = 4.39 → 4 (was 5). Capability = 3 × 0.30 + 1 × 0.25 + 2 × 0.25 + 4 × 0.20 = 2.45 → 2.5 (was 2.7). Other sub-scores, maturity and adoption unchanged.
Board audit (22 Oct 2026, DI 0.4.1)
Decision Index 0.4.1 (22 Oct; reference Jev 1.13 Full 59.22, was 61.26 on 0.4) lists this page’s row at Full 33.73 (#143). Accuracy rule: 8 × 33.73 ÷ 59.22 = 4.56 → 5 (was 4). Capability = 3 × 0.30 + 1 × 0.25 + 2 × 0.25 + 5 × 0.20 = 2.65 → 2.7 (was 2.5). Maturity and adoption unchanged.
