Top pick
Qwen3.7 Flash
$1.70 / month
Best quality per dollar for a steady stream of summaries: quality 54.0 at $1.70 a month, no cited GPQA Diamond score.
Score 46.3 · $0.03 input · $0.13 output per 1M · context 1,000,000 · 12 providers
Job
Cheapest summarization
Which model is the cheapest way to summarize a steady stream of documents, and what will it cost each month?
DRAFT — placeholder editorial, not approved. The score is quality minus a log of the monthly cost, with a cited GPQA Diamond result when the lab published one. The premium card uses raw quality. If the top pick is already the cheapest model, there is no separate budget card.
How this pick is chosen
- Outputs text. Embeddings and image-only models are not in the catalog.
- Lab is one of: openai, anthropic, google, meta, mistral, deepseek, xai, alibaba, moonshotai, zhipuai, cohere, amazon, nvidia, microsoft, ai21.
- Listed by at least 5 providers. A model on only the lab and a reseller does not clear this.
- Not deprecated, and not an alpha release. Beta releases stay in.
- Not a specialized model: OCR, embeddings, rerankers, speech, safeguard, or moderation, matched on the name.
- Context window of at least 32,000 tokens.
- Has a numeric input and output price.
- Reference price is not $0. A zero in Models.dev is often an unpublished price.
- A benchmark counts only when its source is the model's lab (including that lab's Hugging Face org) or the benchmark owner. Artificial Analysis, OpenRouter, and LLM Stats are dropped. When one benchmark has several cited scores, the median is used.
- Summarization score: quality minus 10 times the log of the monthly cost. Doubling the bill costs 10 quality points. Quality starts at 22 and adds reasoning, structured output, tool calls, the log of the context window, and up to 18 from a cited GPQA Diamond score. General-purpose names add 14, efficiency names such as flash or mini add 6, and nano, micro, tiny, or edge names lose 16. A name with no version digit loses 8.
- The top pick is the highest job score. On coding agents, models within 4 points of that leader are treated as a flagship tie: the lower monthly bill is the top pick and the higher bill is the premium pick. Otherwise the premium pick is the best upgrade that costs more than the top pick. The budget pick is the highest job score that costs less than the top pick and sits at or below the 25th cost percentile. Output rates above $50 per 1M tokens stay out of the premium card. Summarization uses raw quality, not quality per dollar, when it chooses that upgrade. Remaining ties break to more providers, then a lower monthly cost, then a newer release date.
Four hundred summaries a day. Each one reads about 3,000 tokens and writes about 400. Nothing is cached.
212 models cleared the bar.303 failed the lab, provider, or specialty bar.3 failed a capability filter.1 listed at $0 were left out.0 had no usable price.
Budget pick
No cheaper model
Qwen3.7 Flash is already the lowest monthly bill that cleared the bar.
Premium pick
GPT-6 Astra
$600 / month
Highest summary quality if you spend more: quality 79.3 before the bill, GPQA Diamond 96 (openai.com).
Quality 79.3 · $10.00 input · $50.00 output per 1M · context 1,050,000 · 32 providers
When not to use it
DRAFT — placeholder. Do not treat this as a reason to avoid the pick. A finished note should say when a fraction of a cent per summary is the wrong trade, for example customer-facing copy or documents in a language the cheap model handles badly.
DRAFT — placeholder. Models listed at $0 are excluded, because a zero in Models.dev is often a missing price rather than a free model.
The three cards follow the inputs. Fit, and who holds each role, stays on the reference workload.
Everyone who cleared the bar, best fit first
| Score | Model | Monthly | Input | Output | Context | Providers |
|---|---|---|---|---|---|---|
| 46.3 | Qwen3.7 Flash | $1.70 | $0.03 | $0.13 | 1,000,000 | 12 |
| 43.1 | DeepSeek V4 Flash | $6.38 | $0.14 | $0.28 | 1,000,000 | 52 |
| 39.8 | DeepSeek V4.1 Flash | $8.28 | $0.15 | $0.60 | 1,000,000 | 64 |
| 37.2 | Qwen3.8 Flash Next | $9.60 | $0.20 | $0.50 | 262,144 | 5 |
| 34.4 | Command R7B | $2.07 | $0.038 | $0.15 | 128,000 | 6 |
| 33.7 | GLM-4.7-FlashX | $4.44 | $0.07 | $0.40 | 200,000 | 9 |
| 33.7 | GPT-5.6 Luna | $12.96 | $0.20 | $1.20 | 1,050,000 | 44 |
| 33.1 | GPT OSS 20B | $3.96 | $0.07 | $0.30 | 131,072 | 22 |
| 32.2 | Qwen3.5 9B | $4.32 | $0.10 | $0.15 | 262,144 | 15 |
| 30.5 | GPT OSS 120B | $8.28 | $0.15 | $0.60 | 131,072 | 44 |
| 29.4 | Gemini 2.5 Flash-Lite | $5.52 | $0.10 | $0.40 | 1,048,576 | 20 |
| 29.3 | Qwen3.5 Flash | $5.52 | $0.10 | $0.40 | 1,000,000 | 9 |