update readme for v1 public beta launch

[Kamilbenkirane]

No description provided.

[claude]

PR Review: Update README for v1 Public Beta Launch

Summary

This PR streamlines the README for the v1 public beta launch, reducing it from a verbose 293 lines to a focused 245 lines (-91, +31). The changes improve clarity and user experience by emphasizing the unified API approach.

---

✅ Strengths

1. Improved Clarity & Focus

• The simplified quick start example (README.md:47-56) is much more effective at demonstrating the "one SDK, every modality" value proposition
• Removing the overly complex multimodal example (old lines 50-91) eliminates confusion for new users
• The new operations table (README.md:95-107) provides clear visibility into available and planned features

2. Better User Experience

• Reordering install instructions to prioritize uv over pip (README.md:193-197) aligns with modern Python tooling preferences
• Removing the verbose "Behavior changes since v0.3.9" section reduces noise for new users
• The simplified "Switch providers" section focuses on the actual value proposition

3. Professional Documentation

• Maintains consistent formatting and structure
• Clear feature matrix with ✓/○/— symbols for status indication
• Beta release callout (line 33) sets appropriate expectations

---

🔍 Issues Found

1. Typo in Quick Start Example ⚠️

Location: README.md:55

embeddings = await celeste.text.embed(["lorep ipsum", "dolor sit amet"], model="gemini-embedding-001")

Issue: "lorep ipsum" should be "lorem ipsum"

Recommendation:

embeddings = await celeste.text.embed(["lorem ipsum", "dolor sit amet"], model="gemini-embedding-001")

2. Inconsistent Empty Lines (Minor)

Location: README.md:13-14, 57-59, 236-237

Multiple consecutive empty lines appear in several places. While not breaking functionality, this is inconsistent with markdown best practices.

Recommendation: Limit to single empty lines between sections for consistency.

3. Missing PR Description

The PR body is empty. For a significant documentation update like this, a brief description of the motivation and key changes would help reviewers and provide better git history.

---

💡 Suggestions for Enhancement

1. Model Name Accuracy

Location: README.md:51

The example uses model="claude-opus-4-5" but based on the "Switch providers" section (line 120), the correct format appears to be "claude-4-5-sonnet". Consider using consistent, real model IDs throughout.

2. Operations Table Clarity

Location: README.md:95-107

Consider adding a brief note about what "○ Planned" means in terms of timeline (e.g., "Coming in Q1 2026" or "Roadmap item"). This helps users understand when they might expect these features.

3. Quick Start Async Context

Location: README.md:47-56

The quick start examples use await but don't show that they need to be in an async context. Consider adding:

import celeste

async def main():
    # One SDK. Every modality. Any provider.
    text = await celeste.text.generate("Explain quantum computing", model="claude-opus-4-5")
    # ...

Or mention that .sync versions are available.

---

📊 Code Quality Assessment

Aspect	Rating	Notes
Clarity	⭐⭐⭐⭐⭐	Significantly improved readability
Accuracy	⭐⭐⭐⭐	One typo, minor model name concerns
Structure	⭐⭐⭐⭐⭐	Well-organized sections
User Experience	⭐⭐⭐⭐⭐	Much easier to understand
Completeness	⭐⭐⭐⭐	Missing PR description

---

✅ Approval Status

Recommendation: Approve with minor fixes

This is a solid improvement to the README that will significantly help users understand the library. The only blocking issue is the typo on line 55 ("lorep ipsum"). Once that's fixed, this is ready to merge.

Required Changes:

• Fix typo: "lorep ipsum" → "lorem ipsum" (line 55)

Optional Improvements:

• Add PR description
• Clean up consecutive empty lines
• Verify model name consistency
• Add async context note to quick start

---

Great work on simplifying the documentation! This will make onboarding much smoother for v1 beta users. 🚀