Top pick
DeepSeek V4 Flash
$1.34 / month
Best cited reasoning score per dollar for everyday chat: GPQA Diamond 88.1 (huggingface.co), $1.34 a month.
Score 84.7 · $0.14 input · $0.28 output per 1M · context 1,000,000 · 52 providers
Pick for Cheapest chat: DeepSeek V4 Flash at $1.34 per month. Best cited reasoning score per dollar for everyday chat: GPQA Diamond 88.1 (huggingface.co), $1.34 a month. Prices from Models.dev. Data last refreshed Oct 16, 2026, 07:15 AM UTC.
Job
Cheapest chat
Which model is the cheapest solid general assistant, and what will it cost each month?
The score is a cited GPQA Diamond result minus a log of the monthly cost. Cards are chosen only from models whose lab published that score. A very small model loses points even when the lab published a number.
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 A zero in Models.dev is often an unpublished price.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.
- General-assistant score: a cited GPQA Diamond result, minus 8 times the log of the monthly cost. A model with no cited GPQA Diamond score loses 40 points before the bill. Active or dense sizes under 8B, and nano, micro, or tiny names, lose 16. Preview and snapshot names lose 6. A name with no version digit loses 8. The cards are chosen only from models with a cited GPQA Diamond score.
- 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.
Two hundred everyday chats a day. Each one is about 800 tokens in and 400 tokens out. Nothing is cached.
214 models cleared the bar.304 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
Nemotron 3.5 Lightning 30B A3B
$0.86 / month
Lower bill with a cited GPQA Diamond result: GPQA Diamond 75.4 (huggingface.co), $0.86 a month.
Score 61.1 · $0.06 input · $0.24 output per 1M · context 262,144 · 13 providers
Premium pick
GPT-6 Astra
$168 / month
Highest cited GPQA Diamond score if you spend more: GPQA Diamond 96 (openai.com), $168 a month.
Score 36.9 · $10.00 input · $50.00 output per 1M · context 1,050,000 · 33 providers
When not to use it
Skip this page for a coding agent, a long document, or an image. GPQA Diamond is a graduate-science quiz, not a measure of tone, instruction following, or your own chat logs.
The reference workload is 200 chats a day, about 800 tokens in and 400 out, with no cache. Change the calculator if your traffic is different. The cards reprice; they do not re-rank.
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
| Caps | ||||||||
|---|---|---|---|---|---|---|---|---|
| 84.7 | DeepSeek V4 Flash | $1.34 | $0.14 | $0.28 | 1,000,000 | 52 | Apr 24, 2026 | |
| 82.8 |  Qwen3.8 Flash Next | $2.16 | $0.20 | $0.50 | 262,144 | 5 | Aug 27, 2026 | |
| 82.0 | DeepSeek V4.1 Flash | $2.16 | $0.15 | $0.60 | 1,000,000 | 67 | Sep 10, 2026 | |
| 76.8 | GPT-5.6 Luna | $3.84 | $0.20 | $1.20 | 1,050,000 | 46 | Jul 9, 2026 | |
| 66.6 |  Muse Glimmer 30B | $4.32 | $0.30 | $1.20 | 131,072 | 13 | Aug 10, 2026 | |
| 63.6 | Qwen3.8 27B | $9.22 | $0.42 | $3.00 | 262,144 | 45 | Aug 14, 2026 | |
| 61.1 | Nemotron 3.5 Lightning 30B A3B | $0.86 | $0.06 | $0.24 | 262,144 | 13 | Aug 11, 2026 | |
| 58.3 |  GLM-5.2 | $17.28 | $1.40 | $4.40 | 1,000,000 | 89 | Jun 13, 2026 | |
| 58.2 | DeepSeek V4 Pro | $15.84 | $1.32 | $3.96 | 1,000,000 | 58 | Apr 24, 2026 | |
| 57.2 | GPT-5.4 mini | $14.40 | $0.75 | $4.50 | 400,000 | 33 | Mar 17, 2026 | |
| 55.9 | Qwen3.8 Max | $24.00 | $2.00 | $6.00 | 1,000,000 | 32 | Aug 3, 2026 | |
| 53.1 | Qwen3.7 Max | $30.00 | $2.50 | $7.50 | 1,000,000 | 31 | May 21, 2026 |