Dashboard
›
LLM
LLM
LLM Model Sammenligning
Samlet overblik over modeller i Hermes-økosystemet — priser, use cases, status og anbefalinger.
9 modeller er gratis lige nu · 25 modeller var gratis sidst de blev testet
15modeller
6free
9paid
4adopted
🆓 Gratis LLM Modeller
OpenRouter :free modeller evalueret med Opus-scenarier. Data fra 2026-05-12. 9 modeller bestod · 25 modeller testet i alt · Se detaljeret free LLM-side for fuld rapport.
| Model | Status | Testdato | Cooldown | Notes |
|---|---|---|---|---|
| qianfan-ocr-fast | Gratis nu | 2026-05-10 | — | Score: 100 · Baidu · 2.025s |
| gemma-4-26b-a4b-it | Gratis nu | 2026-05-10 | — | Score: 100 · Google AI Studio · 0.775s |
| ring-2.6-1t | Gratis nu | 2026-05-10 | — | Score: 100 · Novita · 15.202s |
| nemotron-3-nano-30b-a3b | Gratis nu | 2026-05-10 | — | Score: 100 · Nvidia · 1.032s |
| nemotron-3-super-120b-a12b | Gratis nu | 2026-05-10 | — | Score: 100 · Nvidia · 3.366s |
| nemotron-nano-9b-v2 | Gratis nu | 2026-05-10 | — | Score: 100 · Nvidia · 4.891s |
| gpt-oss-120b | Gratis nu | 2026-05-10 | — | Score: 100 · OpenInference · 23.356s |
| gpt-oss-20b | Gratis nu | 2026-05-10 | — | Score: 100 · OpenInference · 16.383s |
| glm-4.5-air | Gratis nu | 2026-05-10 | — | Score: 100 · Z.AI · 26.493s |
| dolphin-mistral-24b-venice-edition | Rate-limited | 2026-05-12 | 2× 429 | Vent cooldown · Ikke testet endnu |
| gemma-4-31b-it | Rate-limited | 2026-05-12 | 2× 429 | Vent cooldown · Ikke testet endnu |
| llama-3.2-3b-instruct | Rate-limited | 2026-05-12 | 2× 429 | Vent cooldown · Ikke testet endnu |
| llama-3.3-70b-instruct | Rate-limited | 2026-05-10 | 1× 429 | Vent cooldown · Ikke testet endnu |
| minimax-m2.5 | Rate-limited | 2026-05-10 | 1× 429 | Vent cooldown · Ikke testet endnu |
| hermes-3-llama-3.1-405b | Rate-limited | 2026-05-10 | 4× 429 | Vent cooldown · Ikke testet endnu |
| qwen3-coder | Rate-limited | 2026-05-10 | 1× 429 | Vent cooldown · Ikke testet endnu |
| qwen3-next-80b-a3b-instruct | Rate-limited | 2026-05-10 | 1× 429 | Vent cooldown · Ikke testet endnu |
| trinity-large-thinking | Tomt output | 2026-05-12 | — | Svarede ikke · 6.342s |
| cobuddy | Tomt output | 2026-05-10 | — | Svarede ikke · 3.445s |
| lfm-2.5-1.2b-thinking | Tomt output | 2026-05-10 | — | Svarede ikke · 1.23s |
| nemotron-3-nano-omni-30b-a3b-reasoning | Tomt output | 2026-05-10 | — | Svarede ikke · 120.157s |
| nemotron-nano-12b-v2-vl | Tomt output | 2026-05-10 | — | Svarede ikke · 2.683s |
| laguna-m.1 | Tomt output | 2026-05-10 | — | Svarede ikke · 2.78s |
| laguna-xs.2 | Tomt output | 2026-05-10 | — | Svarede ikke · 1.104s |
| lfm-2.5-1.2b-instruct | Andet fail | 2026-05-10 | — | HTTP 200, men check matchede ikke |
✅ 9 bestået
⏳ 8 rate-limited
❌ 7 tomt output
⚠️ 1 andet
💡 Tip: Modeller med "tomt output" kan kræve andre prompter/parametre. Modeller med 429 cooldown kan stadig være gratis — de var bare rate-limited under test.
Se Free LLM-siden for detaljer og retry-plan.
Model Comparison
Filtrer efter type, use case eller søg. Klik på kolonne-overskrifter for at sortere. Priser er per million tokens (input/output).
| Model | Provider | Type | Best for | Price/M tokens | Status | Last checked |
|---|---|---|---|---|---|---|
| Xiaomi MiMo V2.5 Pro | Xiaomi | Paid | reasoninggeneral | $0.40 / $1.20 | adopted | 2026-05-26 |
| DeepSeek V4 Flash | DeepSeek | Paid | codingbudget | $0.14 / $0.28 | useful | 2026-05-26 |
| Qwen 3.6 Plus | Alibaba | Paid | reasoning | $0.50 / $1.50 | useful | 2026-05-25 |
| GPT-5.5 | OpenAI | Paid | generalreasoning | $2.00 / $8.00 | useful | 2026-05-24 |
| Claude Opus 4 | Anthropic | Paid | codingreasoning | $15.00 / $75.00 | useful | 2026-05-24 |
| Gemini 3 Flash | Paid | speedbudget | $0.36 / $1.08 | useful | 2026-05-25 | |
| MiniMax M2.7 | MiniMax | Paid | general | $0.20 / $0.60 | monitoring | 2026-05-20 |
| Llama 4 Scout | Meta | Free | general | Free | useful | 2026-05-25 |
| DeepSeek R1 | DeepSeek | Free | reasoning | Free | useful | 2026-05-25 |
| Qwen 3 30B | Alibaba | Free | general | Free | useful | 2026-05-24 |
| MiMo V2 Pro | Xiaomi | Free | reasoning | Free | useful | 2026-05-25 |
| Mistral Large 3 | Mistral | Paid | general | $2.00 / $6.00 | monitoring | 2026-05-18 |
| Kimi K2.5 | Moonshot | Paid | coding | $0.60 / $1.80 | monitoring | 2026-05-18 |
| Grok 3 | xAI | Paid | general | $3.00 / $15.00 | monitoring | 2026-05-15 |
| Command R+ | Cohere | Paid | RAGtool-use | $2.50 / $10.00 | later | 2026-05-10 |
Ingen modeller matcher din søgning.
Sammenlign to modeller
Vælg to modeller for at se en side-om-side sammenligning.
Free LLM
OpenRouter :free modeller, budget, 429 og coverage.
Åbn FreeDeep Search
Narusya Deep Search forbrug på OpenRouter.
Åbn Deep SearchAnbefalinger
Bedste gratis: MiMo V2.5 Pro. Bedste budget: DeepSeek V4 Flash. Bedste reasoning: Qwen 3.6 Plus.