Weekly Build Brief — 05 October 2026
Dated changes in the LLM API catalogue and GPU rental floor between 2026-10-04 and 2026-10-05: 5 price move(s), 5 GPU rate move(s), 2 added / 3 removed model(s). Read from source APIs, with the monthly bill for a stated workload.
Window: 2026-10-04 → 2026-10-05 (two consecutive dated snapshots of the catalogue, read from the source APIs — never written by a model).
What moved in the API catalogue
| Model | Per 1M tokens | Before | After | Change |
|---|---|---|---|---|
deepseek/deepseek-v4-flash |
output | $0.0560 | $1.2800 | +2185.7% up |
deepseek/deepseek-v4-pro-0813 |
input | $0.2200 | $0.8500 | +286.4% up |
z-ai/glm-5.2 |
output | $3.4900 | $8.0000 | +129.2% up |
deepseek/deepseek-v4.1-flash |
output | $1.2000 | $2.4000 | +100.0% up |
deepseek/deepseek-v4.1-flash |
input | $0.3000 | $0.0030 | -99.0% down |
Full dated log: /llm-costs/changes/.
What it costs on a fixed workload
Reference workload: 20M input tokens + 5M output tokens per month, priced with the observed rates above — recompute it yourself, the rates are the only input.
| Model | Before | After | Monthly difference |
|---|---|---|---|
deepseek/deepseek-v4-flash |
$0.73 | $6.85 | +6.12 $/month |
deepseek/deepseek-v4-pro-0813 |
$29.40 | $42.00 | +12.60 $/month |
z-ai/glm-5.2 |
$19.37 | $41.92 | +22.55 $/month |
deepseek/deepseek-v4.1-flash |
$6.06 | $12.06 | +6.00 $/month |
deepseek/deepseek-v4.1-flash |
$18.00 | $12.06 | -5.94 $/month |
Live rates: /api-pricing/.
GPU rental floor
| GPU | Before | After | Change |
|---|---|---|---|
RTX 5070 Ti |
$0.1156 | $0.1630 | +41.0% |
RTX PRO 6000 Max-Q |
$1.3215 | $0.8022 | -39.3% |
Tesla V100 |
$0.0422 | $0.0556 | +31.8% |
RTX 3090 |
$0.1089 | $0.1347 | +23.7% |
H100 NVL |
$1.9763 | $2.2689 | +14.8% |
Lowest observed hourly price per card: /gpu-pricing/.
Catalogue entries and exits
| Movement | Model |
|---|---|
| added | Comfy-Org/ltx-2 |
| added | microsoft/deberta-v3-small |
| removed | NeoQuasar/Kronos-base |
| removed | cardiffnlp/twitter-xlm-roberta-base-sentiment |
| removed | unsloth/Qwen3.6-27B-GGUF |
2 added, 3 removed between the two snapshots.
Published this week
| Date | Pipeline | Identifier |
|---|---|---|
| 2026-09-28 | openrouter_catalogue | deepseek/deepseek-v4.1-flash |
| 2026-09-28 | openrouter_catalogue | ~deepseek/deepseek-flash-latest |
| 2026-09-28 | openrouter_catalogue | ~z-ai/glm-flash-latest |
| 2026-09-28 | openrouter_catalogue | deepseek/deepseek-v4-flash |
| 2026-09-28 | openrouter_catalogue | deepseek/deepseek-v4-flash |
| 2026-09-28 | openrouter_catalogue | z-ai/glm-5.3-flash |
| 2026-09-28 | openrouter_catalogue | qwen/qwen3.5-35b-a3b |
| 2026-09-28 | openrouter_catalogue | qwen/qwen3-14b |
| 2026-09-28 | openrouter_catalogue | qwen/qwen3-14b |
| 2026-09-28 | openrouter_catalogue | ~deepseek/deepseek-pro-latest |
Every entry is a page on this site; the registry that produces it is data/content_registry.json.
How this page is built
- Prices, GPU rates and catalogue entries are read from the source APIs and stored as dated snapshots (
data/history/pricing/,data/history/gpu/,data/history/models/). - This page compares 2026-10-04 and 2026-10-05 and shows only what crossed the materiality thresholds (5% and $0.02 per 1M tokens for prices, 5% for GPU rates).
- No language model writes this page. The only arithmetic is the monthly bill above, from the rates shown.
- Nothing to report means no page: a week without a material change is not published.
- Observed 2026-10-05 05:55 (local time).
Was this article helpful?
Let us know to improve our AI generation.