SAGE OSS

The open-source model family from SAGEA, designed for community research, fine-tuning, and self-hosted deployments. Fully open weights with permissive licensing.

⚡ Speed

★ Performance

◈ Modalities

Text · Code

Context

16k

Price

USD

$0.05

/M Tokens

$0.15

/M Tokens

Speed

Performance

Context

16k

Price

$0.05 / $0.15 /M Tokens

Features

Try it

https://api.sagea.space/v1

import sagea

client = sagea.Client(api_key="your-api-key")

response = client.chat.completions.create(
    model="sage-oss",
    messages=[{"role": "user", "content": "Hello, SAGE!"}]
)

print(response.choices[0].message.content)

Base URL: https://api.sagea.space/v1 — Auth: Authorization: Bearer $SAGEA_API_KEY

Specifications

Parameters	7B / 14B / 24B
Context Length	16,384 tokens
Max Output	4,096 tokens
Training Data	Up to March 2026
License	Apache 2.0

Other Models

Capabilities

SAGE OSS is the community family (7B / 14B / 24B) with fully open weights under Apache 2.0. Fine-tune with LoRA or full training, self-host, and deploy privately.

Open weights: download and run locally with no restrictions.
Fine-tunable: LoRA + full fine-tuning recipes included.
API-compatible: same POST https://api.sagea.space/v1/chat/completions schema as hosted models, so you can migrate seamlessly.

from transformers import AutoModelForCausalLM, AutoTokenizer
 
model = AutoModelForCausalLM.from_pretrained("sagea/sage-oss-14b")
tokenizer = AutoTokenizer.from_pretrained("sagea/sage-oss-14b")

Use cases

Research, distillation, and domain adaptation
Private / air-gapped deployments
Prototyping before scaling to Actus/Celer in production

Limitations

16k context — use Actus for long-context workloads.
Community support; production SLAs apply to hosted API plans. See Choosing a Model.

SAGE OSS

Features

Try it

Specifications

SAGE 2.4 Actus

SAGE 2.5 Celer

Sonus TTS

Capabilities

Use cases

Limitations

Next steps

On this page