StellarComputerX logoStellarComputerX
Get API Key

Model rankings

Production-ready model rankings.

Compare hosted models by scenario, speed, context, pricing, and availability. Start with a shortlist, then open the registry to test or request a route.

Open model registryCatalog ready

Overall model ranking

#1 scx/llama-3.3-70bMeta · Global apps and multilingual assistant routes
88SCX score

Catalog ready

Live and preview models are grouped so teams can find a usable route faster.

Compare in one table

Context, speed, pricing, status, and best-fit scenarios stay side by side.

Route confidence

Use score bands and dimension leaders to shortlist models before evaluation.

Model intelligence

Model intelligence

Use these charts to see leading routes, coverage balance, and the tradeoffs between cost, latency, and context length.

Top production routesscore
01scx/llama-3.3-70b
88
02scx/qwen2.5-72b-instruct
88
03scx/qwen3-32b
88
04scx/glm-4-32b
87
05scx/qwen2.5-coder-32b
87
06scx/deepseek-r1-32b
86
07scx/deepseek-v3
86
08scx/llama-3.1-8b-instruct
86
Score distribution25 models
090-100
1385-89
680-84
670-79
Provider coverage7
Qwen
5
DeepSeek
3
Mistral
3
BAAI
2
Google
2
Meta
2
01.AI
1
Capability coverage25 models
Chat
10
Reasoning
4
Embedding
3
Vision
3
Code
2
Audio
1
Image
1
Rerank
1
Context envelope25 models
8K4
32K2
64K1
128K14
200K1
1M1
Audio1
Image1
Cost and speed shortlistLow cost / Fast response

Low costscore

01scx/llama-3.1-8b-instruct0.22 Credit / 1K output tokens93
02scx/jina-embeddings-v30.10 Credit / 1K input tokens92
03scx/bge-m30.12 Credit / 1K input tokens91
04scx/e5-mistral-7b-instruct0.18 Credit / 1K input tokens91
05scx/bge-reranker-v2-m30.18 Credit / 1K input tokens89

Fast responsescore

01scx/bge-m3Instant97
02scx/jina-embeddings-v3Instant97
03scx/bge-reranker-v2-m3Very fast93
04scx/e5-mistral-7b-instructVery fast93
05scx/gemma-2-9b-itVery fast93
OverallBest all-around aliases for production trials.ReasoningMulti-step reasoning, agent plans, and tool use.CodeCode generation, review, migration, and copilots.VisionScreenshots, documents, images, and multimodal apps.RetrievalEmbeddings, rerank, RAG, and search quality.ValueLower Credit cost with usable production posture.Low latencyFast response routes for real-time UX.

Ranking view

Overall model ranking

A practical starting point for choosing model aliases across common production workloads.
RankModelProviderBest forContextSpeedPriceStatusScore
01
ME
scx/llama-3.3-70bMultilingual
MetaGlobal apps and multilingual assistant routes128KFast0.88 Credit / 1K output tokensLive
88
02
QW
scx/qwen2.5-72b-instructGeneral chat
QwenCopilots, workflows, customer support128KFast0.95 Credit / 1K output tokensLive
88
03
QW
scx/qwen3-32bReasoning
QwenFast assistants, routing fallback, live UX128KFast0.78 Credit / 1K output tokensLive
88
04
ZH
scx/glm-4-32bGeneral chat
ZhipuCopilots, workflows, customer support128KFast0.74 Credit / 1K output tokensLive
87
05
QW
scx/qwen2.5-coder-32bCoding
QwenCode generation, review, migration plans128KFast0.82 Credit / 1K output tokensLive
87
06
DE
scx/deepseek-r1-32bReasoning
DeepSeekAgent planning, tool use, structured reasoning128KBalanced1.45 Credit / 1K output tokensLive
86
07
DE
scx/deepseek-v3General chat
DeepSeekCopilots, workflows, customer support128KFast0.82 Credit / 1K output tokensLive
86
08
ME
scx/llama-3.1-8b-instructGeneral chat
MetaLow-cost fallback routes and high-volume traffic128KVery fast0.22 Credit / 1K output tokensLive
86
09
GO
scx/gemma-3-27b-itVision
GoogleMultimodal apps, image understanding, assistants128KFast0.58 Credit / 1K output tokensLive
85
10
IN
scx/internlm2.5-20b-chatGeneral chat
InternLMFast assistants, routing fallback, live UX32KVery fast0.36 Credit / 1K output tokensLive
85
11
MI
scx/mixtral-8x22bGeneral chat
MistralCopilots, workflows, customer support64KFast0.78 Credit / 1K output tokensLive
85
12
QW
scx/qwen2.5-vl-72bVision
QwenDocument vision, screenshots, visual reasoning128KBalanced1.05 Credit / 1K output tokensLive
85
13
01
scx/yi-largeMultilingual
01.AIGlobal apps and multilingual assistant routes200KBalanced0.92 Credit / 1K output tokensLive
85
14
OP
scx/internvl3-38bVision
OpenGVLabDocument vision, screenshots, visual reasoning128KBalanced0.96 Credit / 1K output tokensLive
84
15
MI
scx/mistral-large-2Coding
MistralCode generation, review, migration plans128KBalanced1.10 Credit / 1K output tokensLive
83
16
DE
scx/deepseek-r1-70bReasoning
DeepSeekHard reasoning, math, multi-step analysis128KDeep1.95 Credit / 1K output tokensPreview
82
17
GO
scx/gemma-2-9b-itGeneral chat
GoogleLow-cost fallback routes and high-volume traffic8KVery fast0.16 Credit / 1K output tokensLive
81
18
MI
scx/minimax-text-01Long context
MiniMaxLong documents, research packs, policy analysis1MBalanced1.25 Credit / 1K output tokensPreview
81
19
QW
scx/qwen3-235b-a22bReasoning
QwenHard reasoning, math, multi-step analysis128KDeep2.20 Credit / 1K output tokensPreview
81
20
JI
scx/jina-embeddings-v3Embeddings
Jina AIMemory layers, RAG pipelines, long retrieval8KInstant0.10 Credit / 1K input tokensLive
79
21
BA
scx/bge-m3Embeddings
BAAISemantic search, retrieval, knowledge bases8KInstant0.12 Credit / 1K input tokensLive
78
22
MI
scx/e5-mistral-7b-instructEmbeddings
MistralMemory layers, RAG pipelines, long retrieval32KVery fast0.18 Credit / 1K input tokensLive
78
23
BA
scx/bge-reranker-v2-m3Rerank
BAAISearch quality, retrieval reranking, answer grounding8KVery fast0.18 Credit / 1K input tokensLive
76
24
WH
scx/whisper-large-v3Audio
WhisperSpeech transcription and meeting intelligence30 minBalanced0.20 Credit / minutePreview
73
25
ST
scx/stable-diffusion-xlImage
StabilityImage generation, creative workflows, draftsImageBalanced0.65 Credit / imagePreview
71