Open model index

Every open-weights model in the directory — 545 of them, re-pulled and re-ranked every 6 hours (last: 2026-08-16). Run ~80% of the work on one of these, free and private on your own hardware, and escalate only the hard ~20% to a frontier model. Any of them plugs into the local lane.

545 models
Kimi K3-256K
k3
Hosted
— · 1.0M ctx · Elo 1674 · tools · reasoning
$2.00 / $8.00 per 1M · free tier
49 hosts: crof, requesty, aiand
Kimi K3 (EU)
kimi-k3@eu
Hosted
— · 1.0M ctx · Elo 1674 · tools · reasoning
$2.25 / $11 per 1M
1 host: requesty
Kimi K3
umans-kimi-k3
Hosted
— · 1.0M ctx · Elo 1674 · tools · reasoning
$3.00 / $15 per 1M · free tier
2 hosts: umans-ai
GLM-5.2
databricks-glm-5-2
Hosted
— · 1.0M ctx · Elo 1585 · tools · reasoning
$1.40 / $4.40 per 1M
1 host: databricks
GLM-5.2
glm-5-2
Hosted
— · 1.0M ctx · Elo 1585 · tools · reasoning
$0.420 / $1.32 per 1M · free tier
70 hosts: nano-gpt, openrouter, llmgateway
GLM-5.2 (EU)
glm-5-2@eu
Hosted
— · 1.0M ctx · Elo 1585 · tools · reasoning
$1.20 / $4.20 per 1M
1 host: requesty
GLM 5.2 Fast
glm-5p2
Hosted
— · 1.0M ctx · Elo 1585 · tools · reasoning
$1.40 / $4.40 per 1M
1 host: fireworks-ai
GLM-5.2
glm5-2
Hosted
— · 1.0M ctx · Elo 1585 · tools · reasoning
$2.31 / $6.00 per 1M
2 hosts: regolo-ai, wafer.ai
GLM 5.2
umans-glm-5-2
Hosted
— · 405K ctx · Elo 1585 · tools · reasoning
$1.40 / $4.40 per 1M · free tier
2 hosts: umans-ai
GLM 5.2
zai-org-glm-5-2
Hosted
— · 1.0M ctx · Elo 1585 · tools · reasoning
$1.40 / $4.40 per 1M
1 host: venice
DeepSeek V4 Pro (Alibaba Cloud)
alicloud-deepseek-v4-pro
Hosted
— · 1.0M ctx · Elo 1584 · tools · reasoning
$1.69 / $3.38 per 1M
1 host: aihubmix
DeepSeek V4 Pro (DeepSeek)
deep-deepseek-v4-pro
Hosted
— · 1.0M ctx · Elo 1584 · tools · reasoning
$0.478 / $0.956 per 1M
1 host: aihubmix
DeepSeek V4 Pro
v4-pro
Hosted
— · 1.1M ctx · Elo 1584 · tools · reasoning
$0.348 / $0.696 per 1M · free tier
53 hosts: routing-run, crof, ebcloud
DeepSeek V4 Flash (Alibaba Cloud)
alicloud-deepseek-v4-flash
Hosted
— · 1.0M ctx · Elo 1581 · tools · reasoning
$0.140 / $0.280 per 1M
1 host: aihubmix
DeepSeek V4 Flash (DeepSeek)
deep-deepseek-v4-flash
Hosted
— · 1.0M ctx · Elo 1581 · tools · reasoning
$0.154 / $0.308 per 1M
1 host: aihubmix
DeepSeek V4 Flash
umans-deepseek-v4-flash
Hosted
— · 1.0M ctx · Elo 1581 · tools · reasoning
$0.140 / $0.280 per 1M · free tier
2 hosts: umans-ai
DeepSeek V4 Flash
v4-flash
Hosted
— · 1.1M ctx · Elo 1581 · tools · reasoning
$0.063 / $0.125 per 1M · free tier
63 hosts: unorouter, openrouter, opencode-go
Tencent Hy3
hy3
Hosted
— · 262K ctx · Elo 1522 · tools · reasoning
$0.066 / $0.260 per 1M · free tier
15 hosts: nano-gpt, openrouter, kilo
GLM-5.1 (Alibaba Cloud)
alicloud-glm-5-1
Hosted
— · 200K ctx · Elo 1510 · tools · reasoning
$0.840 / $3.38 per 1M
1 host: aihubmix
GLM 5.1
glm-5-1
Hosted
— · 205K ctx · Elo 1510 · tools · reasoning
$0.450 / $2.15 per 1M · free tier
48 hosts: crof, hpc-ai, nano-gpt
GLM 5.1
umans-glm-5-1
Hosted
— · 204K ctx · Elo 1510 · tools · reasoning
$1.40 / $4.40 per 1M · free tier
2 hosts: umans-ai
GLM 5.1
zai-org-glm-5-1
Hosted
— · 200K ctx · Elo 1510 · tools · reasoning
$1.54 / $4.84 per 1M
1 host: venice
Kimi K2.6
k2-6
Hosted
— · 262K ctx · Elo 1509 · tools · reasoning
$0.275 / $1.10 per 1M · free tier
64 hosts: routing-run, llmgateway, vultr
Kimi K2.6 Fast
kimi-k2p6
Hosted
— · 262K ctx · Elo 1509 · tools · reasoning
$0.950 / $4.00 per 1M
1 host: fireworks-ai
MiniMax M3
m3
Hosted
— · 1.0M ctx · Elo 1490 · tools · reasoning
$0.225 / $0.900 per 1M · free tier
36 hosts: empiriolabs, wandb, deepinfra
MiMo V2.5 Pro
mimo-v2-5-pro
Hosted
— · 1.1M ctx · Elo 1474 · tools · reasoning
$0.400 / $0.800 per 1M · free tier
21 hosts: crof, nano-gpt, openrouter
MiMo V2.5 Pro (Crof)
mimo-v2-5-pro-crof
Hosted
— · 1.0M ctx · Elo 1474 · tools · reasoning
$0.400 / $0.800 per 1M
1 host: nano-gpt
Kimi K2.7 Code
kimi-k2-7-code-1100b
Hosted
1100B · 262K ctx · Elo 1473 · tools · reasoning
$0.860 / $3.00 per 1M
1 host: aki-io
Kimi K2.7 Code
databricks-kimi-k2-7-code
Hosted
— · 262K ctx · Elo 1473 · tools · reasoning
$0.950 / $4.00 per 1M
1 host: databricks
Kimi K2.7 Code
kimi-for-coding
Hosted
— · 262K ctx · Elo 1473 · tools · reasoning
self-host · free tier
1 host: kimi-for-coding
Kimi K2.7 Code
kimi-k2-7-code
Hosted
— · 262K ctx · Elo 1473 · tools · reasoning
$0.275 / $1.10 per 1M · free tier
48 hosts: routing-run, crof, inceptron
Kimi K2.7 Code Fast
kimi-k2p7-code
Hosted
— · 262K ctx · Elo 1473 · tools · reasoning
$0.950 / $4.00 per 1M
1 host: fireworks-ai
Kimi K2.7 Code
umans-kimi-k2-7
Hosted
— · 262K ctx · Elo 1473 · tools · reasoning
$0.950 / $4.00 per 1M · free tier
2 hosts: umans-ai
Qwen3.6 Plus
qwen3-6-plus
Hosted
— · 1.0M ctx · Elo 1460 · tools · reasoning
$0.500 / $3.00 per 1M
1 host: togetherai
MiMo V2.5 Thinking
mimo-v2-5
Hosted
— · 1.1M ctx · Elo 1438 · tools · reasoning
$0.140 / $0.280 per 1M · free tier
20 hosts: nano-gpt, impossibl, openrouter
Xiaomi MiMo-V2.5 (free)
xiaomi-mimo-v2-5
Hosted
— · 1.0M ctx · Elo 1438 · tools · reasoning
$0.400 / $2.00 per 1M · free tier
2 hosts: venice, aihubmix
GLM-5
glm-5
Hosted
— · 205K ctx · Elo 1436 · tools · reasoning
$0.480 / $1.90 per 1M · free tier
35 hosts: crof, openrouter, gmicloud
Kimi K2.5
k2-5
Hosted
— · 262K ctx · Elo 1436 · tools · reasoning
$0.350 / $1.70 per 1M · free tier
44 hosts: crof, nano-gpt, llmgateway
Kimi K2.5 (Lightning)
kimi-k2-5-lightning
Hosted
— · 131K ctx · Elo 1436 · tools · reasoning
$1.00 / $3.00 per 1M
1 host: crof
Kimi K2.5
moonshotai-kimi-k2-5
Hosted
— · 262K ctx · Elo 1436 · tools · reasoning
$0.600 / $3.00 per 1M
2 hosts: amazon-bedrock, edenai
GLM 5
zai-org-glm-5
Hosted
— · 198K ctx · Elo 1436 · tools · reasoning
$1.00 / $3.20 per 1M
1 host: venice
GLM-4.7
glm-4-7
Hosted
— · 204K ctx · Elo 1434 · tools · reasoning
$0.200 / $0.800 per 1M · free tier
35 hosts: meganova, nano-gpt, crof
MiMo V2 Pro
mimo-v2-pro
Hosted
— · 1.0M ctx · Elo 1434 · tools · reasoning
$1.00 / $3.00 per 1M · free tier
2 hosts: opencode-go
GLM 4.7
zai-org-glm-4-7
Hosted
— · 198K ctx · Elo 1434 · tools · reasoning
$0.550 / $2.65 per 1M
1 host: venice
Inkling
inkling
Hosted
— · 1.0M ctx · Elo 1405 · tools · reasoning
$0.950 / $4.05 per 1M · free tier
17 hosts: deepinfra, openrouter, edenai
Inkling
inkling-nvfp4
Hosted
— · 1.0M ctx · Elo 1405 · tools · reasoning
$1.20 / $5.00 per 1M
1 host: modal
Inkling (256K)
inkling-peft
Hosted
— · 262K ctx · Elo 1405 · tools · reasoning
$3.74 / $9.36 per 1M
1 host: thinkingmachines
Qwen3.5 397B A17B
qwen3-5-397b
Hosted
397B · 17B active · 262K ctx · Elo 1400 · tools · reasoning
$0.172 / $1.03 per 1M · free tier
29 hosts: orcarouter, merge-gateway, empiriolabs
Qwen3.5 397B-A17B
qwen35-397b
Hosted
397B · 17B active · 262K ctx · Elo 1400 · tools · reasoning
$0.600 / $3.60 per 1M
1 host: llmgateway
MiniMax-M2.7
m2-7
Hosted
— · 262K ctx · Elo 1397 · tools · reasoning
$0.080 / $0.320 per 1M · free tier
34 hosts: llmgateway, empiriolabs, deepinfra
MiniMax M2.7
minimax-m27
Hosted
— · 198K ctx · Elo 1397 · tools · reasoning
$0.375 / $1.50 per 1M
1 host: venice
MiniMax-M2.7
minimax-m2p7
Hosted
— · 196K ctx · Elo 1397 · tools · reasoning
$0.300 / $1.20 per 1M
1 host: fireworks-ai
MiniMax M2.1
m2-1
Hosted
— · 204K ctx · Elo 1387 · tools · reasoning
$0.270 / $1.10 per 1M · free tier
19 hosts: llmgateway, meganova, jiekou
MiniMax-M2.5
minimax-m2-5-230b
Hosted
230B · 196K ctx · Elo 1384 · tools · reasoning
$0.250 / $1.20 per 1M
1 host: aki-io
MiniMax-M2.5
m2-5
Hosted
— · 228K ctx · Elo 1384 · tools · reasoning
$0.220 / $0.880 per 1M · free tier
39 hosts: dinference, openrouter, inceptron
MiniMax M2.5
minimax-m25
Hosted
— · 198K ctx · Elo 1384 · tools · reasoning
$0.270 / $0.950 per 1M
1 host: venice
Gemma 4 31B
gemma-4-31b
Runs local
31B · 262K ctx · Elo 1364 · tools · reasoning
$0.102 / $0.297 per 1M · free tier
32 hosts: llmgateway, crof, openrouter
ollama pull gemma-4:31b
Gemma 4 31B IT
gemma4-31b
Runs local
31B · 262K ctx · Elo 1364 · tools · reasoning
$0.400 / $1.00 per 1M · free tier
5 hosts: nano-gpt, tinfoil, regolo-ai
ollama pull gemma4:31b
Gemma 4 26B A4B IT
gemma-4-26b
Runs local
26B · 4B active · 262K ctx · Elo 1362 · tools · reasoning
$0.050 / $0.290 per 1M · free tier
18 hosts: empiriolabs, cloudflare-workers-ai, cloudflare-ai-gateway
ollama pull gemma-4:26b
Gemma 4 26B A4B IT
gemma4-26b
Runs local
26B · 256K ctx · Elo 1362 · tools · reasoning
$0.100 / $0.500 per 1M
1 host: aki-io
ollama pull gemma4:26b
80/20

How a model gets used

Built-in router

Every request is scored for complexity. Routine work runs on your local model — free, on your hardware, private. Complex work, or work the local model struggles with, escalates to the frontier model you bring a key for. Each escalation is metered against your org budget, so cost can't run away.

Plugging one in

  • Your own machine — install Ollama, pull the model, and the desktop app talks to localhost:11434. Nothing leaves the box.
  • Hosted — point OMNIROUTE_BASE_URL at any OpenAI-compatible gateway; the models above are served by dozens of independent providers, so you're never locked to one.
  • Pick it here — an admin selects the model and every agent in the org routes to it; leave it unset to use the server default.

Ranked by measured LMArena Elo where the boards cover a model (88 of them today), and by how widely a model is hosted otherwise — an honest proxy for what people actually run, never a score we made up.

Directory data from models.dev; quality scores from the LMArena community boards. All product names, logos, and trademarks are the property of their respective owners.