# Open-weight reserve status

**As of:** 2026-09-14
**SoT store:** Scaleway Object Storage `nl-ams` bucket
`railiance-fi-open-weight-reserve` (**live** — FI-WP-0004-T08)
**Transfer:** diskless streaming on Railiance; [operations](../docs/streaming-reserve.md)
**Capacity gate:** **€15 / month** soft (not 850 GiB)

---

## Portfolio snapshot

| Class | Count | Status mix |
| ----- | ----: | ---------- |
| Catalog entries | 15 | + V4-Flash S, + Qwen3.8-27B W/R |
| **verified** on VAULT | 4 | R embeds + Qwen3-8B + R1-Distill-14B |
| **approved**, blobs missing | S giants + Llama-3.2-3B | HF gate / deferred |
| **collecting** | 1 | V4-Flash → Scaleway Glacier, remote stream |
| **candidate** | 4 | W + Qwen3.8-27B |

### Disk / object use

| Location | Role | Approx |
| -------- | ---- | ------ |
| VAULT `models/nomic-ai__nomic-embed-text-v1.5/main` | R embed (staging cache) | ~0.5 GiB verified |
| VAULT `models/BAAI__bge-m3/main` | R embed (staging cache) | ~2.1 GiB verified |
| VAULT `models/Qwen__Qwen3-8B/main` | R instruct (staging cache) | ~15.3 GiB verified |
| VAULT `models/deepseek-ai__DeepSeek-R1-Distill-Qwen-14B/main` | R reason (staging cache) | ~27.5 GiB verified |
| Scaleway bucket | SoT | V4-Flash streaming; final manifest pending |

VAULT ~45 GiB used remains a **cache**, not the reserve.

---

## Decision matrix (2026-09-13)

| Id | Tier | Decision | Blobs |
| -- | ---- | -------- | ----- |
| Qwen3-8B | R | **verified** on VAULT cache | local only |
| Llama-3.2-3B-Instruct | R | **approved** | deferred — HF gate |
| BGE-M3 | R | **verified** on VAULT cache | local only |
| R1-Distill-Qwen-14B | R | **verified** on VAULT cache | local only |
| nomic-embed-text-v1.5 | R | **verified** on VAULT cache | local only |
| **DeepSeek-V4-Flash-0731** | **S1** | **collecting** — **first Scaleway pull** | partial — 166.9 GB / 155.4 GiB source, Glacier |
| Kimi-K3 | S1 secondary | **approved**, deferred until after V4-Flash | no — ~1454 GiB Glacier ≈ €3.69/mo |
| DeepSeek-V3 | S superseded identity | **approved**, deferred | no — V4-Flash is the collectable S1 |
| DeepSeek-R1 full | S2 | **approved**, deferred | no |
| Qwen3-72B | S4 | **approved**, deferred | no |
| Llama-3.3-70B-Instruct | S4 | **approved**, deferred | no |
| Qwen3.8-27B | W/R | **candidate** | no |
| Qwen3-14B | W | candidate | no |
| R1-Distill-32B | W | candidate | no |
| bge-reranker-v2-m3 | W | candidate | no |

**Correction:** there is no `DeepSeek-V4-Flash-0731-12B`. Automated briefs
that used that id were wrong. Do not collect under that name.

---

## Hardware coverage gaps

| Need | Status |
| ---- | ------ |
| Default local instruct (8B) | **Covered** (Qwen3-8B cache) |
| Stronger local instruct (27B) | **Candidate** Qwen3.8-27B |
| Edge micro-agent | Llama-3.2-3B blocked on HF auth |
| Multilingual RAG embed | **Covered** (BGE-M3 + nomic) |
| Local reason | **Covered** (R1-Distill-14B) |
| Frontier-open MIT MoE offline | **Collecting** V4-Flash — remote diskless stream |
| Full open giant (K3) | Cataloged; Glacier-affordable; after V4-Flash |

---

## Next pulls (operator)

1. **FI-WP-0004-T08** bucket live; dedicated IAM key remains outstanding.
2. **FI-WP-0004-T09** monitor the active remote stream; require the complete manifest before marking collected.
3. Set `HF_TOKEN` and pull Llama-3.2-3B-Instruct into `models/` One Zone IA.
4. Decide Qwen3.8-27B vs keeping Qwen3-8B as the R instruct default.
5. Optional: Kimi K3 on Glacier if the month is still under €15.

Tool: `scripts/collect_model.py` (`--s3-bucket` streams without disk staging; explicit `--local-download` for cache only).