LLM Reference

Yi 34B

Released
2023-11-02
Last refreshed
2026-06-01
Status
Researched 3d ago
CodingRAGLong contextClassificationJSON / Tool use

Yi 34B is worth evaluating for coding, rag, and long context when its provider route and context window match the workload.

Use it for

  • Teams evaluating coding, rag, and long context
  • Workloads that can use a 200k context window
  • Buyers comparing 4 tracked provider routes

Do not use it for

  • Vision or document-understanding workloads
Specifications
Released
2023-11-02
Context
200k
Parameters
34B
Architecture
Decoder Only
Knowledge cutoff
2024-03
Specialization
general
Training
finetuned
Created by

Developing AI for creative and generative tasks.

Beijing, China
Founded 2023
Website
Pricing
Output / 1M
$0.380
Input / 1M
$0.250

Cheapest of 5 routes · DeepInfra

About

Yi 34B is 01.AI's Yi (2023/11) model. It offers a 200K-token context window and scores 43 on MMLU PRO.

Yi 34B is 01.AI's 34-billion-parameter bilingual English-Chinese language model. This LLMReference page is aligned to the long-context Yi 34B listing carried in the seed record, with a 200K-token context window, rather than the shorter base card. 01.AI released Yi-34B in November 2023 and also published a Yi-34B-200K extended-context variant for long-document work. The model uses a decoder-only transformer architecture trained independently from scratch; it is similar in shape to Llama-style systems but is not derived from Meta's weights.

The practical reason to consider this row is long bilingual context at a 34B scale. Yi 34B was competitive for its release era on English and Chinese evaluation sets, including a 43 score on MMLU PRO in the current seed data, and it remains useful as a historical baseline for Chinese/English long-context comparisons. It is not the first model to reach for when maximum current reasoning quality matters, but it can still be relevant for teams comparing older open-model deployments or reproducing Yi 1.x results.

Provider coverage is uneven: the seed links Yi 34B to Together AI, Fireworks AI, DeepInfra, Replicate, and Alibaba Cloud, with some listings deprecated or provisioned-only. Treat provider availability and context limits as implementation details to verify before deployment, while this page's model-level summary should be read as the 200K-context Yi 34B entry.

Yi 34B has a 200k-token context window.

Yi 34B input tokens at $0.2/1M, output at $1/1M.

Top use-case fit: coding, agents, and build tasks

Coding

Q/$ B

1 relevant benchmark in the decision map.

RAG

Included by capability and metadata signals in the decision map.

Long context

Included by capability and metadata signals in the decision map.

Provider price ladder

Compare all 5

Compare API pricing across 4 providers for input and output tokens, batch, and cached reads when available.

ProviderInput / 1MOutput / 1MRoute
DeepInfra$0.250$0.380
Serverless
Fireworks AI$0.900$0.900
Provisioned
Replicate API$0.200$1.00
Serverless
Alibaba Cloud PAI-EAS--
ServerlessPartial

Capabilities

Structured Outputs

Benchmark peer barsfor Coding

Benchmark scores(3)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionSource
HumanEval60.0pass@1https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard
Massive Multitask Language Understanding73.25-shothttps://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard
MMLU PRO43.0https://huggingface.co/spaces/TIGER-Lab/MMLU-Pro

Migration checks

No linked migration route is available for this model yet.

Rankings & picks(8)