Weekly Build Brief — 19 October 2026
Dated changes in the LLM API catalogue and GPU rental floor between 2026-10-18 and 2026-10-19: 5 price move(s), 5 GPU rate move(s), 6 added / 7 removed model(s). Read from source APIs, with the monthly bill for a stated workload.
Window: 2026-10-18 → 2026-10-19 (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-pro |
input | $0.2088 | $0.9483 | +354.2% up |
deepseek/deepseek-v4-pro |
output | $0.4176 | $1.8966 | +354.2% up |
deepseek/deepseek-v4.1-flash-1017 |
input | $0.1500 | $0.3000 | +100.0% up |
deepseek/deepseek-v4.1-flash-1017 |
output | $0.6000 | $1.2000 | +100.0% up |
~deepseek/deepseek-pro-latest |
input | $0.1200 | $0.2250 | +87.5% up |
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-pro |
$13.66 | $28.45 | +14.79 $/month |
deepseek/deepseek-v4-pro |
$21.05 | $28.45 | +7.39 $/month |
deepseek/deepseek-v4.1-flash-1017 |
$9.00 | $12.00 | +3.00 $/month |
deepseek/deepseek-v4.1-flash-1017 |
$9.00 | $12.00 | +3.00 $/month |
~deepseek/deepseek-pro-latest |
$42.40 | $44.50 | +2.10 $/month |
Live rates: /api-pricing/.
GPU rental floor
| GPU | Before | After | Change |
|---|---|---|---|
RTX 4090 |
$0.2022 | $0.1215 | -39.9% |
RTX 3090 |
$0.0881 | $0.1213 | +37.7% |
RTX PRO 6000 WS |
$1.0681 | $1.4022 | +31.3% |
RTX 5080 |
$0.1738 | $0.1348 | -22.4% |
RTX PRO 6000 Max-Q |
$1.3215 | $1.0859 | -17.8% |
Lowest observed hourly price per card: /gpu-pricing/.
Catalogue entries and exits
| Movement | Model |
|---|---|
| added | Comfy-Org/Pixal3D |
| added | Harveenchadha/vakyansh-wav2vec2-tamil-tam-250 |
| added | NbAiLab/nb-wav2vec2-1b-nynorsk |
| added | Qwen/Qwen3-VL-8B-Instruct-GGUF |
| added | orcarouter/Qwen3.8-Flash-Next-Uncensored-GGUF |
| added | tencent/Hy-MT2-7B-GGUF |
| removed | QuantTrio/Qwen3.6-35B-A3B-AWQ |
| removed | Qwen/Qwen2.5-7B |
| removed | Xenova/whisper-tiny.en |
| removed | casperhansen/llama-3.3-70b-instruct-awq |
| removed | dbmdz/bert-large-cased-finetuned-conll03-english |
| removed | gabor-hosu/e5-mistral-7b-instruct-bnb-4bit |
| removed | trl-internal-testing/tiny-Qwen3ForCausalLM |
6 added, 7 removed between the two snapshots.
Published this week
| Date | Pipeline | Identifier |
|---|---|---|
| 2026-10-12 | openrouter_catalogue | deepseek/deepseek-v4-flash |
| 2026-10-12 | openrouter_catalogue | deepseek/deepseek-v4-flash-0731 |
| 2026-10-12 | openrouter_catalogue | ~deepseek/deepseek-v4-flash-latest |
| 2026-10-12 | openrouter_catalogue | ~z-ai/glm-flash-latest |
| 2026-10-12 | openrouter_catalogue | ~deepseek/deepseek-flash-latest |
| 2026-10-12 | openrouter_catalogue | ~deepseek/deepseek-flash-latest |
| 2026-10-12 | openrouter_catalogue | z-ai/glm-5.3 |
| 2026-10-12 | openrouter_catalogue | ~z-ai/glm-latest |
| 2026-10-12 | openrouter_catalogue | ~deepseek/deepseek-pro-latest |
| 2026-10-12 | openrouter_catalogue | ~moonshotai/kimi-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 2026-10-18 and 2026-10-19 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 2026-10-19 05:55 (local time).
Was this article helpful?
Let us know to improve our AI generation.