# gemma-4-26B-A4B-it Q4_0 on Ryzen AI Max+ 395 (llama.cpp Vulkan) — needle retrieval success: 95.8 % of requests

On a Beelink GTR9 Pro — AMD Ryzen AI Max+ 395, Radeon 8060S (gfx1151), 128 GB LPDDR5X-8000 unified running llama.cpp (server, Vulkan backend), b9049 (server_fingerprint b9049-2496f9c14) with ggml-org/gemma-4-26B-A4B-it-GGUF @ bb4531cda34d (Q4_0), the measured needle retrieval success was 95.8 % of requests (share of all needle requests). The measurement was taken under the frozen long-context-v1@2026-08-03 workload; evidence level lab_repeated, 3 valid runs across 1 physical unit. The value is re-derived from the raw run records on every CI build.

## Statement

> Share of requests where the model retrieved the embedded fact across a 2k-32k-token ladder, EN and RU, default operating mode.

## Evidence

| Field | Value |
| --- | --- |
| Claim id | `strix.gemma4.longctx.c1.needle-success` |
| Value | 95.8 % of requests |
| Metric | needle retrieval success |
| Aggregation | share of all needle requests |
| System | Beelink GTR9 Pro — AMD Ryzen AI Max+ 395, Radeon 8060S (gfx1151), 128 GB LPDDR5X-8000 unified |
| Runtime | llama.cpp (server, Vulkan backend), b9049 (server_fingerprint b9049-2496f9c14) |
| Model artifact | ggml-org/gemma-4-26B-A4B-it-GGUF @ bb4531cda34d (Q4_0) |
| Workload scope | long-context-v1@2026-08-03 |
| Evidence level | lab_repeated |
| Status | active |
| Runs | 3 across 1 unit(s) |
| Published | 2026-08-03 |
| Last changed | 2026-08-03 |

**Limitations.** Three repeated runs on one unit. Default operating mode of the model (reasoning emitted); disabling reasoning via the chat template was not tested for this model family.

## Raw evidence

- Sealed run bundle `run-20260803-gm-lc-a`: https://github.com/botAGI/agmind-lab/tree/main/runs/run-20260803-gm-lc-a
- Sealed run bundle `run-20260803-gm-lc-a-r2`: https://github.com/botAGI/agmind-lab/tree/main/runs/run-20260803-gm-lc-a-r2
- Sealed run bundle `run-20260803-gm-lc-a-r3`: https://github.com/botAGI/agmind-lab/tree/main/runs/run-20260803-gm-lc-a-r3
- Derivation query: https://github.com/botAGI/agmind-lab/blob/main/catalog/claims/sql/needle-success.sql
- Machine-readable: https://agmind.ai/claims/strix.gemma4.longctx.c1.needle-success.json · full registry https://agmind.ai/claims.json

## Cited on

- https://agmind.ai/reports/do-local-llms-invent-answers/
- https://agmind.ai/reports/model-quant-backend-strix-halo/
- https://agmind.ai/measured/gemma-4-26b-a4b-it-q4-0-on-ryzen-ai-max-395/

## Cite this

AGmind Systems Lab. gemma-4-26B-A4B-it Q4_0 on Ryzen AI Max+ 395 (llama.cpp Vulkan): needle retrieval success — 95.8 % of requests (share of all needle requests; 3 runs on 1 unit; evidence level lab_repeated; workload long-context-v1@2026-08-03). Claim strix.gemma4.longctx.c1.needle-success. https://agmind.ai/claims/strix.gemma4.longctx.c1.needle-success/

Registry changes feed: https://agmind.ai/claims/changes.xml · Corrections to published results: https://agmind.ai/errata/ · Data licensed CC BY 4.0.
Source page: https://agmind.ai/claims/strix.gemma4.longctx.c1.needle-success/