Flux 2 Dev
bfl/flux-2/devFlux 2 Dev by BFL - generate stunning images from text prompts with state-of-the-art AI. Supports multiple aspect ratios, styles, and high-resolution output for creative and commercial use.
- Base price
- $0.0012USD / run
- Execution
- async
- Model type
- image
- Input fields
- 6
Try the model
Playground
Your output will appear here
Complete the inputs, then click Run.
Specifications
Pricing
- Base price
- $0.0012 / run
Context & modalities
- Input
- Schema-defined
- Output
- image
Capabilities
- Chat
- Not supported
- Vision
- Not supported
- Reasoning
- Not supported
- Structured output
- Not supported
- Function calling
- Not supported
- Audio input
- Not supported
Access
- Provider
- BFL
- Model ID
- bfl/flux-2/dev
- Execution
- async
- API
- Unified Run API
- Endpoint
- /v1/run
API README
FLUX.2 [dev]
FLUX.2 [dev] is a 32-billion-parameter rectified-flow transformer for image generation from text instructions. The model is positioned for maximum-quality FLUX.2 work when latency is not the primary constraint, making this endpoint a strong choice for deliberate concept development and detailed final-image exploration.
Its official model card describes leading open text-to-image performance and an ability to follow instructions that can also drive image generation, editing, and combination across the underlying model. This route focuses that visual intelligence on text-to-image output and returns one generated image asynchronously.
Highlights
- High-end text-to-image generation. FLUX.2 [dev] is identified as the maximum-quality option in the FLUX.2 family when generation quality matters more than latency.
- Detailed instruction interpretation. Translate a layered written brief into an image while retaining subject, environment, composition, lighting, material, and stylistic direction.
- Character, object, and style understanding. The underlying model is trained to represent these visual identities without task-specific fine-tuning, strengthening descriptive prompts that depend on a coherent design language.
- Prompt-upsampling gains. FLUX.2 [dev] benefits significantly from expanded prompts, so a concise idea can be developed into a richer production brief before generation.
Pricing
| Unit | Base price |
|---|---|
| Per request | $0.0012 |
When to Use
| Use case | Why this model fits |
|---|---|
| High-detail concept art | Prioritize the development model's maximum-quality positioning over low-latency output. |
| Layered cinematic scenes | Express subject, action, setting, camera, lighting, and texture in one descriptive brief. |
| Character and object design | Build a coherent visual identity from detailed appearance, material, and style language. |
| Prompt-expansion workflows | Turn a short concept into richer production prose before sending the final request. |
| Controlled visual comparisons | Hold the seed stable while comparing meaningful prompt revisions. |
Prompt Guide
Use descriptive prose rather than a keyword pile. Establish the main subject and framing first, then add action, environment, lighting, lens or perspective, palette, materials, and fine surface detail. If expanding a short idea, remove contradictions and repeated adjectives before generation.
{
"prompt": "Dutch-angle close-up of a cheerful survivor in a post-apocalyptic market, dust on the face, layered weathered fabrics, harsh late-afternoon sunlight and deep shadows, desaturated earth palette, tactile material detail, cinematic gritty realism",
"guidance_scale": 2.5,
"num_inference_steps": 28,
"aspect_ratio": "16:9",
"output_format": "png",
"seed": 41827
}
Technical Specs
| Property | Details |
|---|---|
| Model ID | bfl/flux-2/dev |
| Required input | prompt |
| Optional inputs | guidance_scale, num_inference_steps, aspect_ratio, output_format, seed |
| Guidance scale | 0–20; default 2.5 |
| Inference steps | 4–50; default 28 |
| Aspect ratios | 21:9, 16:9 (default), 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 |
| Output formats | png (default) or jpeg |
| Output | Downloadable image URL with optional content type |
| Execution | Asynchronous |
| License | FLUX.2-dev Non-Commercial License |
Related Models
bfl/flux-2/dev/edit— use FLUX.2 [dev] for image editing.bfl/flux-2/kontext/dev— use the Kontext development route for context-aware edits.bfl/flux-1.1/pro— compare with a production-oriented FLUX generation model.
Start building
Send your first request
OpenAI-compatible endpoint with unified authentication and usage tracking.
# The quoted heredoc keeps Unicode and shell metacharacters unchanged.
result=$(curl --fail-with-body --silent \
-X POST "https://api.sandbase.ai/v1/run" \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
-H "Content-Type: application/json" \
--data-binary @- <<'SANDBASE_JSON'
{
"model": "bfl/flux-2/dev",
"prompt": "Dutch angle close-up of a survivor in post-apocalyptic setting, dust-covered face, dramatic harsh sunlight creating deep shadows, happy expression, desaturated dystopian color palette, gritty realism",
"aspect_ratio": "16:9",
"output_format": "png",
"guidance_scale": 2.5,
"num_inference_steps": 28
}
SANDBASE_JSON
)
run_id=$(printf '%s' "$result" | jq -r .id)
for attempt in $(seq 1 120); do
status=$(printf '%s' "$result" | jq -r .status)
case "$status" in completed|failed|timeout) break ;; esac
sleep 2
result=$(curl --fail-with-body --silent \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
"https://api.sandbase.ai/v1/run/$run_id")
done
status=$(printf '%s' "$result" | jq -r .status)
[ "$status" = completed ] || { echo "Generation ended: $status" >&2; exit 1; }
printf '%s\n' "$result"Choose your model
Compare models
| Model | Context | Input / 1M | Output / 1M | Released |
|---|---|---|---|---|
Flux 2 DevThis model BFL | — | — | — | Nov 23, 2025 |
BFL | — | — | — | May 28, 2026 |
BFL | — | — | — | May 21, 2026 |
BFL | — | — | — | May 14, 2026 |
BFL | — | — | — | Mar 17, 2026 |
BFL | — | — | — | Mar 17, 2026 |