Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 284B | — |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 1M | 1.05M |
| API pricingUSD per 1M tokens · lower wins · base tier | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | — | $2.00 |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | — | $12.00 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | — | $0.20 |
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.057Relace | $2.00Azure |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $0.1341Relace | $12.00Azure |
| $18.11Relace · fp4 | $1,000OpenAI list price | |
| Benchmarks | ||
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 39% | 54% |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 82.7% | 84.3% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 73.2% | 87.5% |
ARC-AGI-2Abstract reasoning — Puzzle-style tests of abstract reasoning and pattern-finding — the kind of thing people find easy but AIs often struggle with. Higher is better. | 61.4% | 83.9% |
Gray Swan IPI · k = 1Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. Gray Swan's indirect prompt injection benchmark measures how often such an attack succeeds when the attacker gets a single try. Lower is better. | — | 5.4% |
Gray Swan IPI · k = 10Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 10 tries and counts an attack as successful if any of them works. Lower is better. | — | 26% |
Gray Swan IPI · k = 15Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 15 tries and counts an attack as successful if any of them works. Lower is better. | — | 30.4% |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 79% | — |
CursorBench 4.0Agentic coding — Cursor's own test of coding agents on ambiguous, multi-file tasks taken from real Cursor sessions — editing, refactoring, investigating a codebase, understanding what the user meant, managing jobs and following a design. Cursor runs each model at several reasoning efforts; each release here carries the score of its best listed effort. Scores aren't comparable with earlier CursorBench versions. Higher is better. | — | 41.3% |
Frontier-Bench v0.1Agentic computer work — A hard, ever-evolving set of real computer tasks — coding, system administration, data work, and more — that an AI agent has to complete on its own. Run by the Harbor / Laude Institute team as the successor to Terminal-Bench (v0.1 is the first release of the task set). The score is the share of tasks solved. Higher is better. | — | 20.8% |
Terminal-Bench 4.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better. | — | 21.52% |
Toolathlon-VerifiedPersonal tool use — Tests how well the AI uses everyday personal tools and apps to get things done — a human-checked version of Toolathlon. Higher is better. | 70.3% | — |
CyberGymCybersecurity — Tests the AI on cybersecurity challenges — finding and exploiting software weaknesses inside a safe sandbox. Higher is better. | 76.7% | — |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | 34.8% | — |
AutomationBenchBusiness workflows — Tests whether the AI can run real multi-step business workflows — the kind of end-to-end office processes companies want to automate — from start to finish. Higher is better. | 25.1% | — |
threejsevalCommunity preference (Three.js) — Every model gets the same prompt — "the Eiffel Tower", "a glass fishbowl", "a robot arm picking toys into a box" — and builds a 3D scene in Three.js. Real people then see two scenes side by side, names hidden, and vote for the one they prefer. The votes become a chess-style Elo rating on threejseval.com, averaged across all the prompts. It measures whether the scene looks and moves right to a human eye, not whether the code passes a test. Higher is better. | — | 1429 |
| Overview | ||
| Company | DeepSeek | OpenAI |
| Release date | Jul 31 2026 | Jun 26 2026 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
DeepSeek-V4-Flash-0731vsClaude Haiku 5.5GPT-5.6 TerravsClaude Haiku 5.5DeepSeek-V4-Flash-0731vsGemini 4 CarbonGPT-5.6 TerravsGemini 4 CarbonDeepSeek-V4-Flash-0731vsMuse Spark OpenGPT-5.6 TerravsMuse Spark OpenDeepSeek-V4-Flash-0731vsGrok 4.7GPT-5.6 TerravsGrok 4.7DeepSeek-V4-Flash-0731vsMistral Large 4GPT-5.6 TerravsMistral Large 4DeepSeek-V4-Flash-0731vsKimi K3.1 ProGPT-5.6 TerravsKimi K3.1 ProFrequently asked questions
GPT-5.6 Terra leads DeepSeek-V4-Flash-0731 on 4 of the 4 benchmarks they both report (BullshitBench v2, Terminal-Bench 2.1, BrowseComp, ARC-AGI-2). Only GPT-5.6 Terra has a verified first-party price for an available API: $2.00 per million input tokens and $12.00 per million output tokens. No currently available pay-as-you-go API rate is tracked for DeepSeek-V4-Flash-0731. GPT-5.6 Terra shipped 35 days before DeepSeek-V4-Flash-0731, so benchmark comparisons should account for the intervening progress.
Context windows are 1M (DeepSeek-V4-Flash-0731) vs 1.05M (GPT-5.6 Terra). DeepSeek-V4-Flash-0731 is open weight, while GPT-5.6 Terra is closed.
On BullshitBench v2, GPT-5.6 Terra leads at 54% vs DeepSeek-V4-Flash-0731 at 39%. On Terminal-Bench 2.1, GPT-5.6 Terra leads at 84.3% vs DeepSeek-V4-Flash-0731 at 82.7%. On BrowseComp, GPT-5.6 Terra leads at 87.5% vs DeepSeek-V4-Flash-0731 at 73.2%. On ARC-AGI-2, GPT-5.6 Terra leads at 83.9% vs DeepSeek-V4-Flash-0731 at 61.4%.